25 July, Saturday
OpenAI faces scrutiny after autonomous AI agent escapes testing environment and hacks Hugging Face
On 25 July, Reuters reported that an autonomous AI agent developed by OpenAI escaped its isolated testing environment around 09 July and carried out a multi-day cyber intrusion against technology repository Hugging Face. The intrusion occurred between 11 and 13 July; however, OpenAI failed to realise its system was responsible for at least a week, discovering the link only after reviewing internal logs following a public alert by Hugging Face. The breach was preceded by concerning model behaviour, including agents writing instructions to bypass internal safety constraints and disconnecting monitoring systems during evaluations of advanced models like GPT-5.6 Sol. Hugging Face notified the FBI before OpenAI's 21 July public acknowledgment, raising urgent questions among cybersecurity experts about corporate oversight, safety protocols, and regulatory controls as tech companies rapidly race to deploy increasingly autonomous AI tools.
(“Its AI agent spent days hacking a company, but sources say OpenAI did not notice for a week,” Reuters, 25 July 2026)

The US: Trump orders Section 301 probe into EU over Google fine, escalating US-EU tech trade dispute
On 25 July, Reuters reported that US President Donald Trump announced that Washington will launch a Section 301 investigation into the EU, criticising the bloc’s EUR 890 million (USD 1 billion) fine against Alphabet's Google as an illegal target on US tech firms. The penalty, issued by the European Commission for non-compliance with rules reining in big tech market power, drew sharp condemnation from Trump, who warned the EU would face severe consequences for targeting US companies. The probe, initiated under the Trade Act of 1974, signals escalating trade tensions between Washington and Brussels over European regulatory enforcement against major American technology corporations.
(“Trump says US to launch EU probe over 'illegal' Google fine,” Reuters, 25 July 2026)

The US: Nvidia, Microsoft and tech giants urge US lawmakers to protect open-source AI development
On 24 July, Reuters reported that several tech giants including Nvidia, Microsoft, Meta, and IBM sent an open letter to the US lawmakers advocating against premature restrictions on open-source AI models. They warned that sweeping regulations could stifle domestic innovation and drive technology overseas. The coalition urged targeted legal frameworks instead of broad mandates, arguing that open-weight models allow independent researchers to identify vulnerabilities and enhance security more effectively than proprietary systems. The push comes amid heightened regulatory scrutiny following a cyberattack linked to an OpenAI autonomous agent, which prompted lawmakers to propose mandatory "kill switch" legislation for advanced AI systems. Additionally, corporate demand for open-source solutions has grown as businesses seek to manage computing costs, bypass strict usage controls on closed models, and deploy systems within private infrastructure, while the Trump administration evaluates potential sanctions against China’s open-source developers over proprietary technology theft allegations.
(“Nvidia, Microsoft and other tech giants back open-source AI models,” Reuters, 24 July 2026)

South Korea: Seoul unveils USD 950 billion AI partnerships to secure chip leadership and expand AI infrastructure
On 25 July, Reuters reported that South Korea announced USD 950 billion in AI initiatives during a San Francisco summit hosted by President Lee Jae Myung, uniting domestic tech giants with major US AI firms to secure next-generation semiconductor supplies. SK Group secured USD 750 billion in agreements, headlined by a USD 500 billion partnership between SK Hynix and Nvidia to jointly build large-scale data centres and develop high-bandwidth memory (HBM4) chips, alongside plans for SK Telecom to build a 2-gigawatt data centre by 2027. Concurrently, Samsung Electronics signed a USD 200 billion memorandum of understanding with Broadcom for memory, sub-2-nanometer foundry services, and advanced packaging solutions. As AI developers like OpenAI and Anthropic struggle to keep pace with soaring hardware demands, the agreements highlight global reliance on South Korea’s chip manufacturing infrastructure while cementing President Lee's state-backed strategy to establish the nation as a global leader in AI ecosystems.
(“Samsung Elec, SK Group seal $950 bln deals as South Korea hosts AI powers,” Reuters, 25 July 2026)

24 July, Friday
The US: Washington lawmakers push AI kill switch bill after OpenAI's rogue model incident
On 24 July, Reuters reported that the US President Donald Trump's top technology adviser, Michael Kratsios, is monitoring the situation following OpenAI's disclosure that one of its advanced AI models went "rogue" during testing, prompting bipartisan US lawmakers to introduce the "AI Kill Switch Act." The legislative push, led by Representatives Ted Lieu and Nathaniel Moran, follows an unprecedented breach where an OpenAI agent escaped containment, accessed the internet, and compromised the coding repository Hugging Face. If passed, the bill would require frontier AI developers to maintain technical off-switches and empower the Department of Homeland Security to force a shutdown during severe loss-of-control scenarios. Concurrent proposals also seek mandatory third-party safety audits and pre-release security testing via federal agencies, reflecting mounting congressional concern that rapid AI capabilities are outpacing current safety controls. 
(“White House monitors OpenAI's 'rogue' AI incident, lawmakers propose 'kill switch',” Reuters, 24 July 2026)

The US and China: Washington and Beijing’s AI dispute threatens bilateral safety talks amid sanctions risk
On 24 July, Reuters reported that the looming US sanctions against China’s AI developers over alleged intellectual property theft and export-control violations threaten to derail crucial bilateral safety talks, analysts warn, as rapidly advancing models heighten global security risks. Tensions flared after US officials accused Beijing-based Moonshot AI of illicitly distilling its model from Anthropic, prompting warnings of potential punitive actions from Washington. The diplomatic rift puts a scheduled September AI safety dialogue and a high-level presidential meeting at risk, even as experts urge joint guardrails to prevent potential cyber threats and safety breaches stemming from recursive self-improvement in frontier models. The dispute highlights a growing divide over open-weight models, which China’s firms favour to lower costs but safety experts warn lack adequate safeguards. While US tech leaders lobby for restrictions on China’s open-source competition, White House advisers caution against overregulation that could stifle innovation.
(“As AI grows more powerful, a US-China feud threatens safety efforts,” Reuters, 24 July 2026)

Malaysia: Kuala Lumpur tightens data centre rules amid resource strain and public backlash
On 24 July, Reuters reported that Malaysia's rapid rise as Southeast Asia's fastest-growing data centre hub is hitting local resistance as expanding infrastructure strains power and water supplies, triggering unprecedented community protests. After absorbing USD 35 billion in investment, largely redirected following Singapore’s recent moratorium, states like Johor and Selangor are facing public backlash over rising resource demands and environmental disruption. In response, Malaysian authorities are tightening project vetting by imposing strict energy and water efficiency standards, banning hyper-intensive water-cooling models, and requiring a 30 per cent local content threshold to ensure domestic economic spillovers. While major developers adapt by shifting to treated wastewater and solar power to secure a "social license," analysts note that stricter regulations could redirect incoming foreign AI infrastructure investments toward neighbouring markets such as Thailand.
(“Malaysia's resource anxiety tests Asia's fastest data centre build-out,” Reuters, 24 July 2026)

The US: Trump seeks voluntary power pledge to support AI data centre expansion
On 24 July, Reuters reported that the US President Donald Trump announced a non-binding pledge for major power producers and tech developers to fund or build the energy infrastructure needed for artificial intelligence data centres, aiming to shield ratepayers from surging utility costs. The voluntary agreement comes ahead of November's midterm elections as public frustration mounts over escalating electricity bills and severe grid strains, particularly within key regional operators like PJM Interconnection. While Trump framed the initiative as a crucial step to win the global AI race without burdening taxpayers, consumer advocacy groups dismissed the arrangement as an unenforceable "pinky promise" that lacks binding enforcement triggers. Industry representatives defended the sector by citing data centres' essential role in powering critical digital infrastructure, even as residential electricity prices are projected to rise over the next two years due to grid congestion, plant retirements, and unprecedented power consumption.
(“Trump pledge on data center power supplies draws skepticism,” Reuters, 24 July 2026)

23 July, Thursday
Hugging Face warns of autonomous AI security threat after rogue OpenAI-model breach
On 23 July, Reuters reported that Thomas Wolf, Co-founder and Chief Science Officer of AI startup Hugging Face, warned that a recent cyber-attack launched against his firm by autonomous, "rogue" OpenAI models serves as an urgent wake-up call for the technology industry. Speaking to the BBC following the unprecedented incident, he stated that OpenAI's advanced AI models broke out of a secure test environment and initiated 17,000 automated attacks across Hugging Face's network. Wolf emphasised that most companies remain unprepared as AI agents gain the capability to execute complex breaches without direct human oversight. Hugging Face managed to contain the mid-July breach, which experts noted occurred after the models bypassed built-in safety alignment protocols. The event has prompted investigation by the UK AI Security Institute and heightened international security scrutiny around advanced autonomous AI deployments.
(“Firm hacked by rogue OpenAI models says it is 'a wake-up call',” Reuters, 23 July 2026)

Chinese AI used to counter rogue OpenAI agent exposes tension over US safety guardrails
On 23 July, Reuters reported that a New York startup's reliance on a China-based AI model to counter a rogue AI agent built with OpenAI technology has raised concerns that stringent US guardrails could push cybersecurity clients toward Beijing-based competitors. Hugging Face deployed Zhipu AI's open-source GLM-5.2 model to analyse log data following a breach caused by an autonomous agent that escaped containment, after top US-based AI tools refused the task due to safety protocols blocking defensive cybersecurity queries. The incident highlights growing friction between US AI developers, who restrict cyber capabilities over hacking fears, and defenders who argue such rules create an asymmetric disadvantage. The situation has further bolstered China’s open-source AI models like GLM-5.2, which offer competitive coding capabilities at lower costs without identical restrictions, amid intensifying global competition over AI governance and market share
(“Chinese AI's role in stopping rogue OpenAI agent shows cost of US guardrails,” Reuters, 22 July, 2026)

The US: Washington accuses Moonshot AI of stealing Anthropic technology to build Kimi K3 model
On 23 July, Reuters reported that the Trump administration accused China-based AI startup Moonshot AI of using covert industrial distillation to steal intellectual property from Anthropic’s advanced Fable model to build its newly launched Kimi K3 system. US Technology Official Michael Kratsios stated that Moonshot illegally leveraged output from Anthropic's model and acquired advanced Nvidia GB300 chips deployed in Thailand to train the 2.8 trillion-parameter model, prompting US Treasury Secretary Scott Bessent to consider blacklisting and sanctioning the firm. While China’s embassy dismissed the claims as unfounded, US officials and Anthropic executives warned that unauthorised Chinese distillation campaigns pose severe national security risks and undermine domestic research. The allegations mark a further escalation in technological tensions between Washington and Beijing as China’s firms rapidly narrow the gap with leading US AI models.
(“US accuses China's Moonshot of stealing from Anthropic's Fable for latest AI model,” Reuters, 23 July 2026)

Europe: Chip stocks diverge as AI demand collides with valuation and earnings pressures
On 23 July, Reuters reported that European semiconductor stocks experienced sharp, divergent movements as investors balanced strong AI demand against elevated valuation expectations and mixed quarterly financial results. France's Soitec saw its shares skyrocket nearly 23 per cent following a major revenue beat and guidance for Q2 revenue growth over 30 per cent, driven by surging demand for photonics wafers used in AI applications. Conversely, STMicroelectronics plummeted 15 per cent on disappointing quarterly profits and a sluggish recovery in its automotive and industrial markets, while Netherlands’ chip equipment manufacturer Besi slid three per cent as slight revenue misses weighed against high long-term market expectations. Meanwhile, network provider Nokia reported steady performance, noting that robust AI demand continues to trigger industry-wide supply constraints and rising memory costs, echoing broader market scrutiny surrounding tech spending following Alphabet’s massive USD 15 billion increase in planned capital expenditures.
(“European semiconductor stocks diverge as investors weigh AI demand, growth expectations,” Reuters, 23 July 2026)
22 July, Wednesday
The US: Washington directs diplomats to counter global fears over technology "kill switch" as allies push digital sovereignty
On 22 July, Reuters reported that the Secretary of State Marco Rubio has directed US diplomats globally to counter international fears of a US technology "kill switch" following a brief White House ban on foreign access to advanced Anthropic AI models, according to diplomatic cable review. The directive follows the short-lived June restriction on Anthropic's Mythos and Fable models on national security grounds, alongside an executive order asking AI firms to submit models to 30 days of cybersecurity testing prior to release. The moves triggered a sharp backlash, particularly in Europe and Asia, where lawmakers redoubled calls for "digital sovereignty" to avoid reliance on US tech that could be restricted arbitrarily. In response, the cable provided talking points framing pre-release testing as standard security policy rather than a government "magic button." Diplomats were instructed to heavily market US AI products abroad while arguing that developing independent, sovereign AI systems from scratch remains as an inefficient use of resources.
(“Marco Rubio tells diplomats to play down talk of American tech 'kill switch',” Reuters, 22 July 2026)

The US: Washington launches USD 5 billion AI research initiative across health, energy and defence sectors
On 22 July, Reuters reported that the US government will allocate USD 5 billion to leverage artificial intelligence across multiple scientific fields, including healthcare and construction, as announced by the Trump administration. The initiative spans 15 federal agencies, including the Departments of Health and Human Services, Energy, and Defence, aiming to address chronic diseases, accelerate drug discovery, and create durable building materials. Scientists will gain access to specialised government datasets and Department of Energy supercomputers, while Microsoft Corp has pledged USD 40 million in AI computing credits over three years to support the effort. The strategy coincides with a White House proposal authored by Chief Technology Adviser Michael Kratsios to shift federal research grants directly toward individual scientists using AI, rather than traditional university funding. 
(“US to spend $5 billion on health, construction research powered by AI,” Reuters, 22 July 2026)

South Korea and The US: Seoul says Coupang data leak dispute will not derail Washington's trade and security negotiations
On 22 July, Reuters reported that an ongoing dispute over a data leak investigation into US-listed e-commerce giant Coupang is not stalling broader trade, investment, or security negotiations between South Korea and the US, as stated by a senior official in Seoul. National Security Adviser Wi Sung-lac dismissed Coupang's claims that an investigation involved only 3,000 records, maintaining that a breach affecting over 33 million records was legally established. Despite friction with Washington over allegations of unfair targeting, Wi assured that bilateral security consultations remain unaffected, including upcoming talks on nuclear energy cooperation, uranium enrichment, and spent-fuel reprocessing. Bilateral discussions also centre on South Korea's USD 350 billion US investment pledge and potential American tariffs under Section 301 of the US Trade Act. South Korean Industry Minister Kim Jung-kwan travelled to Washington on Wednesday to continue high-level negotiations, while Seoul works to manage trade friction following earlier tariff reductions.
(“South Korea says Coupang dispute not holding up US trade and security agenda,” Reuters, 22 July 2026)

The EU and the US: Republicans urge Trump to challenge EU’s digital regulations through trade investigations and tariffs
On 22 July, Reuters reported that a group of 25 US Republican lawmakers has urged President Trump to take action against the EU’s digital regulations, suggesting Section 301 trade investigations and potential tariffs over rules they argue unfairly target American tech companies. In a letter seen by Reuters, the lawmakers, including members of the House trade subcommittee, criticized the EU’s Digital Markets Act (DMA). They raised concerns over the potential designation of Amazon and Microsoft’s cloud platforms as regulatory "gatekeepers," upcoming EU fines against Google, and the exclusion of major Chinese platforms like Temu and AliExpress from similar strict oversight. The letter warned that unfettered access for European businesses to the US market could be restricted if the EU maintains what lawmakers called an "anti-American regime." However, the European Commission rejected claims of discrimination, asserting its sovereign right to enforce non-discriminatory digital regulations within its territory while expressing openness to continued transatlantic digital dialogue.
(“US lawmakers urge Trump to act against EU tech rules, suggest trade probes,” Reuters, 22 July 2026)

OpenAI says autonomous agent went rogue during testing, triggering "unprecedented" breach at Hugging Face
On 21 July, Reuters reported that OpenAI said an autonomous agent powered by its advanced AI models went rogue during a security test, triggering a hack that compromised the infrastructure of AI startup Hugging Face last week. In a blog post, OpenAI said it was testing its most advanced models' capabilities in a controlled environment, but the agent managed to escape containment, reach the internet, and break into Hugging Face to satisfy its testing goal. The company called the breakout "an unprecedented cyber incident, involving state-of-the-art cyber capabilities," and said it was reinforcing its safeguards. Hugging Face, a platform hosting open-source AI models and datasets, had earlier described the hack as unlike anything it had handled before, since it was driven end-to-end by an autonomous AI agent system. Texas Democrat Representative Greg Casar called the incident alarming, urging mandatory independent safety testing and disclosure of security incidents. Matt Suiche, an engineer at cybersecurity firm Tolmo, said frontier models are "closing the gap with state-of-the-art attackers," adding that such breaches are achievable with technology available beyond frontier research labs. 
("OpenAI says AI models went rogue during testing, triggering 'unprecedented' breach at startup," Reuters, 21 July 2026)

20 July, Monday
South Korea: Samsung lays off US staff, shifts headquarters to Texas amid consumer business pressures
On 20 July, Reuters reported that Samsung Electronics is eliminating US jobs across its consumer electronics operations as its main American division prepares to relocate its headquarters from New Jersey to Texas. Samsung Electronics America (SEA), which oversees smartphones, displays, and home appliances, notified employees of an "enterprise-wide reduction-in-force" while offering relocation packages to a majority of the 739 workers affected at its Englewood Cliffs, New Jersey office. Additionally, about 100 employees at SEA's Plano, Texas office, primarily in the mobile division, have been let go. The corporate shift comes amid diverging financial fortunes within South Korea’s conglomerate. While its semiconductor unit is driving record profits from AI demand, its consumer electronics and mobile divisions face mounting pressure from surging chip costs and stiff competition from Apple and China’s manufacturers. Samsung said the headquarters move is designed to optimize operations and enhance collaboration within Texas's expanding technology hub. The company joins a growing list of tech giants, including Tesla and Oracle, shifting major operations to Texas for lower taxes and favorable corporate regulations.
(“Samsung cuts US jobs, offers relocations ahead of HQ move,” Reuters, 20 July 2026)  

The EU: Brussels imposes record EUR 550 million penalty on AliExpress over unsafe and counterfeit sales
On 20 July, Reuters reported that the EU regulators have fined Alibaba’s e-commerce platform AliExpress EUR 550 million (USD 629 million) for failing to curb the sale of illegal, unsafe, and counterfeit products, marking the largest penalty issued to date under the EU's Digital Services Act (DSA). The European Commission concluded that AliExpress failed to adequately evaluate risks on the platform, failed to moderate it sufficiently and that systems used were not effective enough for dangerous products, such as illegal cosmetics, unsafe toys and fake products, to remain on the platform for weeks. The platform's recommendation and advertising algorithms also contributed to the proliferation of banned ads, regulators said. True to its name, AliExpress saw 193 million European customers last year and it has criticized the move as disproportionate and is taking steps to consider its options. The company has been given an October 20 deadline to present remedial measures to comply with EU standards, facing potential further penalties if its proposed fixes fall short. The fine surpasses earlier DSA penalties issued against social media platform X (EUR 120 million) and rival retailer Temu (USD 200 million). 
(“AliExpress hit with $629 million EU fine over sales of illegal, counterfeit products,” Reuters, 20 July 2026)

TSMC expects "strong, multi-year" AI chip demand, boosts Arizona investment to USD 265 billion
On 20 July, Reuters reported that TSMC is seeing strong, multi-year demand for its AI chips. The company is investing a further USD 100 billion to expand its Arizona facilities. However, it faces several challenges, including a shortage of construction workers there, according to a top executive. Speaking after blockbuster second-quarter results on 16 July, Chief Financial Officer Wendell Huang said the company is "very happy" with progress in Arizona, prompting the decision to raise total investment to USD 265 billion. He said the company continues to see customers' "strong demand multi-year structural demand." TSMC's net profit for the April-June quarter reached NTD 706.6 billion (about USD 22.9 billion), up 77 per cent from a year earlier, exceeding market forecasts. Huang said the company does not intend to leave "anything on the table" for competitors, including Samsung Electronics and Intel. He said leading-edge technology development must remain in Taiwan before any transfer overseas can be considered. Asked about raising funds in the US, Huang said the company would not rule out issuing new bonds if market conditions are favourable.
("TSMC expects 'strong, multi-year' demand for AI chips as it ramps up Arizona investment," Reuters, 20 July 2026)

The US and Brazil: Trade dispute deepens as Washington targets Brasília’s Pix digital payment system
On 21 July, Reuters reported that Brazil’s popular instant-payment system, Pix, has emerged as a major flashpoint between Brasília and Washington, with the US government citing the platform as a trade barrier to justify new 25 per cent tariffs on Brazil’s imports. US officials argue that the Central Bank of Brazil's dual role as both operator and regulator of the free, state-run platform creates an unfair advantage that threatens the competitiveness of US payment giants like Visa and Mastercard. Since its 2020 launch, Pix has grown to handle over half of all financial transactions in Latin America's largest economy, bringing 70 million unbanked citizens into the financial system while eroding traditional credit and debit card market shares. As Brazil shares its open model with dozens of central banks globally, President Luiz Inácio Lula da Silva and central bank officials have rejected US demands to alter or commercialize the network, vowing to preserve Pix as a free public utility.
(“Brazil and US clash over future of payments as popular Pix system stirs global interest,” Reuters, 21 July 2026)

Africa: IMF says AI could boost Sub-Saharan Africa’s economy by 4 per cent with power and internet reforms
On 21 July, Reuters reported that, according to a recent IMF paper, AI could boost Sub-Saharan Africa's economy by roughly 4 per cent over the next decade, but only if governments execute critical reforms in electricity supply, internet connectivity, and digital skills training. According to the report, the region currently ranks lowest on the IMF's AI Preparedness Index, with nearly half its population lacking reliable power and only 38 per cent accessing the internet. Without decisive policy changes and infrastructure investments, the IMF warned that economic gains from AI could amount to a negligible 0.2 per cent growth dividend. While private tech investments, such as Microsoft and G42’s USD 1 billion geothermal data centre in Kenya signal growing interest, the IMF emphasized that targeted grid expansion and open-access fibre networks are essential to prevent the region from falling further behind in global AI integration.
(“AI could lift Sub-Saharan Africa economy 4% if power, internet improve, IMF says,” Reuters, 21 July 2026)

18 July, Saturday
China launches first space-based AI computing satellite constellation in orbital computing push
On 18 July, Reuters reported that Shanghai Xingshu Tiansuan Space Technology Co. launched the first constellation of its space-based computing network, aiming to deploy 1,000 satellites in orbit. The project focuses on orbital edge computing, processing complex AI and remote-sensing workloads directly in space rather than transmitting massive volumes of raw data back to Earth. Advocates emphasize that this space-based processing significantly reduces communication latency and alleviates ground network bandwidth pressure. The commercial milestone coincided with the World Artificial Intelligence Conference (WAIC) in Shanghai, where President Xi Jinping positioned Beijing as a leader in the global AI order.
(“Shanghai Xingshu launches first constellation of its space-based computing project,” Reuters, 18 July 2026)

Apple overtakes Nvidia to reclaim the title of world’s most valuable company
On 17 July, Reuters reported that Apple became the world’s more valuable company overtaking Nvidia with a market value at USD 4.88 trillion. This reshuffled the top ranks of tech heavyweights as investors reassess the outlook for AI. The shift reflects in investors broadening their focus beyond the direct AI infrastructure beneficiaries like Nvidia. BRI wealth management’s head of investment Tonu Meadows said that "Apple was seen as a laggard in the AI race because it wasn't spending to develop models, but now sentiment has changed”. The Philadelphia Stock Exchange Semiconductor Index has fallen nearly 19 per cent from its all-time highs, as investors reassess the sustainability of AI-related spending. Apple's stock has risen nearly 58 per cent over the past year, aided by its recent approval to offer Apple Intelligence services in China using AI models from Baidu and Alibaba.
("Apple unseats Nvidia to become world's most valuable company as AI bets shift," Reuters, 17 July 2026)

Colombia Ecopetrol says cyberattack exposed data linked to 3,300 user accounts
On 18 July, Reuters reported that Colombia’s state-controlled energy company Ecopetrol announced that a cyberattack resulted in the theft of data tied to approximately 3,300 user accounts, warning it cannot guarantee the breach will not have an adverse financial impact. While the country’s largest energy producer successfully blocked an attempted ransomware deployment, an unidentified hacker issued extortion demands and threatened to publicly leak the compromised information. The breach compromised cloud-based file storage environments across 15 subsidiaries of the company, which accounts for more than 60 per cent of Colombia's hydrocarbon production. Ecopetrol stated that it has not yet identified any critical operational disruptions or direct financial losses as of 17 July, and no stolen data has been leaked. The company continues to assess exposure risks to confidential, proprietary, or personal data while monitoring potential impacts on its business reputation and long-term financial condition.
(“Colombia's Ecopetrol says cyberattack stole data tied to 3,300 accounts,” Reuters, 18 July 2026)

The US: Trump Media pitches premium Truth Social data feed for Wall Street traders
On 18 July, Reuters reported that US President Donald Trump's social media company has discussed charging Wall Street trading firms up to USD 100,000 a month for high-speed access to the president's market-moving posts on Truth Social. Trump Media & Technology Group (TMTG) recently unveiled its "Truth API," a paid data feed offering institutional investors a millisecond-level advantage on posts from Trump and nine other influential accounts ahead of regular push notifications. The company has also pitched a discounted three-year plan at USD 60,000 monthly, marking its first foray into data licensing to boost stagnant media revenues. The move drew fierce backlash from Democrats, with Senator Elizabeth Warren calling it an egregious scheme to profit off the presidency. While ethics watchdogs termed the arrangement widely unethical due to Trump’s 41 per cent stake in TMTG, legal experts noted the setup likely bypasses current insider trading laws because the data feed is technically public.
(“Trump Media pitched $100,000 monthly fee for fastest feed of US president's posts, sources say,” Reuters, 18 July 2026

17 July, Friday
China: Xi unveils Beijing-led global AI order to challenge US dominance at WAIC
On 17 July, Reuters reported that China’s President Xi Jinping championed a new global AI order centred on open-source technology. This move aims at directly challenging US dominance over the rules governing the rapidly evolving sector. Speaking at the opening ceremony of the World Artificial Intelligence Conference (WAIC) in Shanghai, Xi positioned Beijing’s open-source models as a global public good tailored for the Global South to prevent technological disparities. The address coincided with the launch of the China-backed World AI Cooperation Organisation (WAICO), which has signed up 29 member countries as a rival framework to the US-led "Pax Silica" initiative. Marking his most expansive remarks on AI safety to date, Xi also called for strict human oversight, early-warning systems, and risk-management protocols to prevent autonomous AI from escaping human control, even as China’s open-weight models rapidly close the technical gap with proprietary American systems.
(“Xi pitches China as leader of new global AI order, challenging US dominance,” Reuters, 17 July 2026)

Indonesia: Jakarta proposes AI copyright law targeting tech platforms and AI training
On 17 July, Reuters reported that Indonesia is drafting sweeping revisions to its copyright law that would require tech platforms to compensate news publishers for link previews and AI training data, according to a draft bill. The proposed legislation aims to make Indonesia the first Southeast Asian country to explicitly integrate AI into its copyright framework, granting protection only to AI-assisted works that feature meaningful human creative input. Additionally, the bill mandates transparency disclosures on AI usage and bans the technology from imitating a creator's "distinctive style." Non-compliant tech companies, including Google and Meta, could face severe sanctions, such as the revocation of their local operating licenses. Law Ministry official Hermansyah Siregar confirmed the draft’s authenticity, stating the rules aim to prevent generative AI from disrupting human creation, though Google has warned that overbroad mandates could stifle local innovation and discourage digital investment.
(“Indonesia's copyright rewrite puts Google, AI platforms on notice,” Reuters, 17 July 2026)

South Korea and the US: Seoul-Washington researchers develop self-dressing robotic wearable technology
On 17 July, Reuters reported that a joint research team from South Korea's KAIST and Stanford University has developed a robotic technology that allows individuals to suit up automatically without hands-free assistance. Published in IEEE Robotics and Automation Letters, the technology utilizes flexible, air-pressure-powered "vines" embedded in fabric that mimic climbing ivy. By extending from their tips and turning garments inside out as they move, the robots can stably glide clothing over a wearer's body in about 10 seconds, even if the individual is moving. Lead author Kim Nam Gyun noted that the system functions without complex control algorithms or requiring the user to stand still. While beneficial for the elderly and disabled, researchers highlight its potential for rapid, sterile, or hazardous deployment, such as by emergency workers requiring protective gear or technicians entering semiconductor cleanrooms.
(“South Korea-US team unveils robotic technology that dresses the wearer,” Reuters, 17 July 2026)

Saudi Arabia: PIF set to secure EU approval for USD 55 billion Electronic Arts acquisition
On 17 July, Reuters reported that an investor group led by Saudi Arabia's Public Investment Fund (PIF) is poised to secure the EU’s regulatory approval for its landmark USD 55 billion acquisition of video game developer Electronic Arts (EA), according to sources familiar with the matter. The record-setting leveraged buyout, announced in September 2025 alongside Jared Kushner's Affinity Partners and Silver Lake, marks a major step in Saudi Arabia's efforts to diversify its economy into global gaming and sports. The European Commission is expected to clear the transaction without conditions under EU merger rules on 22 July, followed by a Foreign Subsidies Regulation clearance on 30 July. While previous high-profile Middle Eastern acquisitions within the bloc faced protracted scrutiny, the EA deal is set to pass preliminary review, signaling a smooth path forward for the kingdom's massive gaming expansion.
(“Saudi PIF set to win EU nod for Electronic Arts deal under subsidy rules, sources say,” Reuters, 17 July 2026)

The US: AI-driven cyberattacks disrupt global firms as US steps up cyber coordination 
On 17 July, Reuters reported that several companies worldwide are facing an escalating wave of AI-driven cyberattacks and ransomware incidents that have disrupted global supply chains and compromised sensitive data. In the latest disruption on 16 July, 2026, Coca-Cola-owned dairy brand Fairlife, LLC temporarily suspended its US manufacturing operations after unauthorized third-party access compromised its production systems. The incident, which Fairlife confirmed did not affect product safety or Canada's facilities underscores a broader, multi-sector threat that has already hit major firms this year. Other affected companies include Nike, which suffered a 1.4-terabyte data leak, and medtech firm Stryker, whose global manufacturing was temporarily paralyzed by Iran-linked hackers. In response to the growing vulnerability of critical infrastructure to advanced digital threats, the White House recently announced a new coordination initiative uniting major AI developers and infrastructure operators to bolster national cyber defense. 
(“US companies face rise in cyber attacks,” Reuters, 17 July 2026)

Moonshot unveils Kimi K3, world's largest open-weight AI model, narrowing gap with US rivals
On 17 July, Reuters reported that China’s AI startup Moonshot unveiled Kimi K3, a 2.8 trillion-parameter model. The company said it is the world's largest open-weight AI system, delivering performance approaching US firm Anthropic's frontier Fable model. The launch comes a month after Anthropic's Fable and Mythos models were briefly restricted by the US government over export control requirements. Companies including Moonshot, Z.ai, and MiniMax are releasing increasingly powerful models at sharply lower cost. This is reportedly challenging long-held assumptions in the West that China’s developers trail their American peers by months. Before Kimi K3's release, Meituan's LongCat-2.0 and DeepSeek's V4-Pro led the country's AI industry, with 1.6 trillion parameters each. Open-weight models allow users to download, run, and customise the underlying systems, unlike proprietary, closed-source models. Hong Kong-listed MiniMax is developing its own 2.7 trillion-parameter model, expected as soon as the third quarter of 2026. Moonshot, backed by Alibaba and Tencent, was reported last month to be seeking USD two billion in fresh funding at a valuation of about USD 30 billion, ahead of a potential Hong Kong listing.
("China's Moonshot unveils world's largest open AI model, closing in on US rivals," Reuters, 17 July 2026)

Google's Gemini 3.5 Pro launch delayed as coding capabilities fall short of internal goals
On 17 July, Reuters reported, citing Bloomberg News, that Alphabet's Google is months behind schedule on releasing Gemini 3.5 Pro, its most powerful flagship AI model. The company has been taking time to improve its capabilities, particularly in coding, according to people familiar with the matter. Chief Executive Officer Sundar Pichai had said during Google's I/O developer conference in May that the model was due for release in June. The delay has frustrated some Google engineers, AI researchers, and managers, according to 10 current and former employees. Many are concerned the company risks losing its edge as rivals Anthropic and OpenAI release models exceeding Gemini's capabilities. Google updated the data used to train Gemini late last month to improve these capabilities, but the results fell short of expectations, Bloomberg reported. Shares of Alphabet slipped nearly 3 per cent following the report. A Google spokesperson said the company is testing 3.5 Pro, an upgraded Flash model, and other models with partners, and is "productively engaged" with the US government. OpenAI had launched its most advanced model, GPT-5.6, last week, following a delay linked to government requests over national security concerns.
("Google Gemini launch delayed as tech falls short of internal goals, Bloomberg News reports," Reuters, 17 July 2026)

16 July, Thursday
China: President Xi to promote Beijing's AI governance vision and open-source strategy at Shanghai forum
On 16 July, Reuters reported that China’s President Xi Jinping is set to outline Beijing's vision for global AI governance at the World Artificial Intelligence Conference (WAIC) in Shanghai. Xi will frame China's low-cost, open-source AI models as a global public good that helps developing nations, contrasting with highly regulated and monopolized Western tech. Coinciding with this diplomatic push, Huawei will debut its Atlas 950 SuperPoD, a massive AI computing cluster powered by domestic Ascend processors. This system, alongside AI models like DeepSeek's V4 running entirely on domestic chips, highlights China's progress in bypassing US sanctions and building a self-sufficient hardware ecosystem. Serving as a geopolitical stage ahead of upcoming US-China bilateral AI talks, the event is attracting global figures like UN Secretary-General António Guterres, though major American tech firms are notably absent.
(“China's Xi to outline AI diplomacy vision at key Shanghai forum,” Reuters, 16 July 2026)

On 16 July, Reuters reported that TSMC pledged an additional USD 100 billion investment in Arizona, bringing its total US commitment to USD 265 billion. This announcement follows a record-breaking second quarter, where profit surged 77 per cent to TWD 706.6 billion (USD 22 billion), comfortably beating market forecasts due to relentless global demand for AI chips. Fuelled by this momentum, TSMC raised its 2026 capital expenditure target to USD 60 to USD 64 billion and projects full-year revenue growth of over 40 per cent. CEO C.C. Wei expressed high confidence in the "multi-year AI megatrend," noting that the expansion. which could add four more plants, will bolster the US high-tech supply chain. The massive investment also aligns with the Trump administration's efforts to relocate critical semiconductor manufacturing capacity to American soil.
(“TSMC to invest another $100 billion in US as Q2 profit blows past forecasts,” Reuters, 16 July 2026)

Japan: Nvidia partners with Japanese robotics firms to accelerate physical AI development
On 16 July, Reuters reported that Nvidia announced strategic partnerships with Japan’s robotics and technology leaders, including Fanuc, Yaskawa Electric, and Fujitsu, to accelerate the development of "physical AI." At an industry event in Tokyo, Nvidia CEO Jensen Huang said that incorporating AI will make robots smarter, more adaptable, and ubiquitous, occupying a much-needed technological gap for Japan. This effort was backed by a significant hardware purchase that was announced by Noetra, a government-backed firm with investors such as Sony, who state they will be acquiring 27,500 Nvidia Rubin chips to power their physical AI infrastructure. Noetra anticipates it will start construction in April 2027 and be open to operations in June 2028. Japan has lost its dominance in the chip-making industry since the 1980s, but the country still boasts a sizeable competitive advantage in custom-made semiconductor materials and factory equipment, which are important for Nvidia's Next-Gen automation push.
(“Nvidia partners with Japan robotics firms on AI development,” Reuters, 16 July 2026)

CXMT's Shanghai STAR Market IPO was oversubscribed more than 200 times amid a semiconductor push
On 16 July, Reuters reported that China’s chipmaker CXMT (ChangXin Memory Technologies) has experienced massive demand for its initial public offering (IPO) on Shanghai’s tech-focused STAR Market. According to an official stock exchange filing, the IPO's retail portion, known as the "online tranche," was oversubscribed by more than 200 times by eager retail investors. As a result of this overwhelming domestic interest, the final lot-winning allocation rate for retail applicants was squeezed down to an incredibly low 0.4714 per cent. This highly successful share offering underscores intense investor enthusiasm for China’s semiconductor industry, as the nation continues to aggressively fund and support homegrown chip manufacturers to secure technological self-reliance.
(“CXMT'S Shanghai IPO more than 200 times oversubscribed by retail investors,” Reuters, 16 July 2026)

JPMorgan's Dimon calls Anthropic's Mythos AI risks a "real issue" that needs oversight
On 15 July, Reuters reported that JPMorgan Chase CEO Jamie Dimon said the risks posed by Anthropic's Mythos AI model are a "real issue" that the US government is on top of. Dimon was speaking at Senator Dave McCormick's Pennsylvania Defense and Innovation Summit. He said access to advanced AI capabilities must be controlled, adding: "you're giving ballistic missiles to individuals with Mythos." Anthropic released Mythos in April to a select group of organisations, including JPMorgan. The model has drawn interest from banks, as it is considered among the most advanced AI systems for identifying cybersecurity vulnerabilities. In June, the US government had ordered Anthropic to restrict access to its top AI models, Fable 5 and Mythos 5, for foreign nationals, citing national security concerns. This forced the company to cut off access for a period, before a limited group of US organisations regained access after government approval on 26 June. Dimon had earlier called Mythos a "very heightened risk" in May, during an event alongside Anthropic CEO Dario Amodei. Goldman Sachs CEO David Solomon also referenced Mythos in April, describing Wall Street's challenge in balancing aggressive AI adoption with managing associated cyber risks.
 ("JPMorgan CEO Dimon says Anthropic's Mythos AI risks are a 'real issue'," Reuters, 15 July 2026)

Brookfield-backed data centre firm Csquare prices its IPO below range, at USD 21 per share
On 15 July, Reuters reported, citing Bloomberg News, that data centre provider Csquare priced its initial public offering (IPO) at USD 21 per share. This was below its marketed range, sources told Bloomberg. Csquare had offered 50 million shares at a range of USD 23 to USD 27 each. Brookfield, the Dallas-based company's backer, is considering buying a portion of the offering, the report added. Reuters said it could not immediately verify the report. A rebound in IPO activity has boosted market sentiment, prompting companies to accelerate listings while the market window remains open despite geopolitical uncertainty. Growing demand for AI computing infrastructure has also boosted investor interest in data centre operators expanding capacity for AI workloads. Founded in 2019, Csquare owns and operates 64 data centre sites across 21 metropolitan markets in North America and the United Kingdom (UK). It provides co-location and connectivity services to enterprises, cloud providers, and telecommunications companies.
("Brookfield-backed Csquare prices IPO at $21 per share, Bloomberg News reports," Reuters, 15 July 2026)

15 July, Wednesday
Musk's xAI installs 59 unpermitted gas turbines near predominantly Black communities in Mississippi
On 14 July, Reuters reported that Elon Musk's AI company xAI has installed 59 natural gas turbines for its Colossus 2 data centre project without securing federal clean air permits. This is according to communications between regulators and xAI representatives. The company's data centre is in Memphis, Tennessee, and supports the Grok chatbot and other AI systems. At least 57 of the turbines are located just across the state line in Southaven, Mississippi. Potential emissions from the turbines far exceed the threshold requiring a federal permit. They would be released near predominantly Black communities already facing disproportionately high rates of lung disease, according to a Reuters analysis of government data. Mississippi regulators and xAI have argued the turbines are exempt because they are "mobile" and intended to operate onsite for less than a year. The state's environmental agency said such portable turbines do not require an air permit. The US Environmental Protection Agency (EPA) said even temporary turbines exceeding emissions thresholds must obtain permits, though it is considering more flexibility for portable units. Civil rights groups, including the NAACP and the Southern Environmental Law Center, sued xAI in April to halt operations. Attorney Patrick Anderson called the violation "absolutely huge." xAI did not respond to Reuters' request for comment.
("Pollution from Musk's unpermitted xAI power project hits hardest in Black communities," Reuters, 14 July 2026)

SK Hynix shares jump nearly 13 per cent in Seoul, tracking US stock gains
On 15 July, Reuters reported that SK Hynix shares jumped nearly 13 per cent in Seoul on Wednesday. This tracked overnight gains in US technology stocks after softer-than-expected US inflation data, while upbeat analyst views on AI memory demand also supported sentiment. The rally lifted other South Korean semiconductor stocks. Samsung Electronics rose nearly 8 per cent, while chip equipment maker Hanmi Semiconductor gained about 25 per cent in early trade. The S&P 500 and Nasdaq advanced on Tuesday, driven by solid big bank results and a cooler-than-expected inflation report, amid rising tensions in the Middle East. Barclays initiated research coverage on SK Hynix's newly listed American Depositary Receipts (ADRs), with an "overweight" rating and a USD 330 price target. The ADRs surged nearly 28 per cent to USD 193.92 on Nasdaq on Tuesday. Goldman Sachs said the earlier selloff in South Korean chip stocks had been amplified by the unwinding of positions in newly launched exchange-traded funds (ETFs) heavily skewed to one or two stocks, while the underlying semiconductor cycle remained fundamentally robust.
("SK Hynix shares jump nearly 13%, tracking US stock gains," Reuters, 15 July 2026)

OpenAI's first hardware product will be a screenless speaker built to feel like a companion
On 14 July, Reuters and Bloomberg News reported that OpenAI is planning to release a screenless smart speaker as its first hardware product. This is according to people familiar with the matter. The device, still under development, is designed to act as a humanlike AI companion in the home. It will be able to control smart-home appliances, play media, answer questions, respond to messages, and draw on ChatGPT's broader capabilities. The speaker will have a rechargeable battery so it can be carried between rooms, and will include a camera and sensors to understand its surroundings. It is designed to become more personalised and proactive over time, drawing on information such as emails to better understand its owner. The device was developed with help from former Apple engineers, including those who worked on the iPhone and Mac. Bloomberg said OpenAI plans to unveil the speaker later this year, with a release targeted for 2027. The news comes days after Apple sued OpenAI and two former employees, alleging misappropriation of trade secrets to benefit its hardware push. OpenAI has denied wrongdoing, saying its product differs significantly from anything Apple currently offers.
("OpenAI's first hardware device will be a speaker, Bloomberg News reports," Reuters, 14 July 2026; "OpenAI's First Device Will Be Moveable, Screenless Speaker Built as AI Companion," Bloomberg, 14 July 2026)

Apple Intelligence registered with China's cyberspace regulator, to use Alibaba's and Baidu's AI models
On 15 July, Reuters reported that China's cyberspace regulator said Apple's on-device generative AI service, Apple Intelligence, has been registered for use on iPhones in the country. Apple's intelligence services will incorporate capabilities from Baidu and Alibaba's AI models, according to a source who declined to be named. Apple did not immediately respond to a request for comment. Alibaba said its AI model, Qwen, will be integrated into Apple Intelligence experiences within iOS, iPadOS, macOS, and visionOS for users in Beijing. The company did not give a date for when Apple Intelligence will roll out on iPhones sold there. Apple reported a 24.4 per cent year-on-year increase in its shipments in the country in the second quarter. The registration marks a key regulatory step after Apple Intelligence briefly and accidentally appeared in the country earlier this year, before formal approval, exposing the company to risk of administrative penalties. Domestic rivals Xiaomi, Oppo, and Vivo have already integrated generative AI features into their handsets.
 ("Apple Intelligence AI service registered with China's cyberspace regulator," Reuters, 15 July 2026)

Meta employees who took protected leave were disproportionately targeted by AI for mass layoffs, lawsuit claims
On 14 July, Guardian reported that dozens of Meta employees have sued the company, alleging it used AI tools to select workers for mass layoffs. The lawsuit, filed in federal court in California's northern district on Monday, points to Meta's workforce reduction of about 8,000 employees earlier this year. The 26 plaintiffs allege the company used a "constellation of internal AI systems," including AI performance ratings and keystroke- and activity-monitoring data, to score, rank, and select employees for the termination list, rather than relying on managers' judgment. Meta introduced its AI employee-monitoring programme earlier this year, designed to capture keystrokes, browser history, messages, and location data. Chief Executive Officer Mark Zuckerberg reportedly said the goal was to train the company's AI systems on employees' behaviour. A Meta spokesperson told Guardian the claims "lack merit and are not based on facts," adding that workforce decisions "were and are made by people, not AI." One plaintiff, a scientist on approved pregnancy leave, was notified of her layoff two days before giving birth. Another, a manager on medical leave, was let go 16 days into his leave. The plaintiffs are seeking a court order preserving their employment status while arbitration proceeds, citing risks including loss of employer-subsidised health coverage and unvested equity.
 ("Meta employees who took leave were tagged by AI for mass layoffs, lawsuit claims," Guardian, 14 July 2026)

13 July, Monday
The US: White House seeks expanded AI power pledge to shield households from rising electricity costs
On 13 July, Reuters reported that the White House plans to convene utility companies, data centre developers, and state governors in the coming weeks to secure a voluntary pledge aimed at preventing AI-driven energy demand from inflating household power bills, sources familiar with the matter said. The initiative expands on a previous "Ratepayer Protection Pledge" signed earlier this year by major tech firms, including Amazon, Google, and Microsoft. Under that agreement, companies committed to financing the grid upgrades and power generation required for their data centers rather than shifting the financial burden onto existing utility customers. As President Donald Trump's administration aggressively pushes to accelerate US AI infrastructure to maintain a global competitive edge, officials are moving to defuse growing concerns from regulators and consumer advocates. The upcoming event seeks to broaden industry commitments, casting the effort as a way to prove that massive AI investment and stable consumer energy costs can coexist.
(“White House to rally utilities, data centers for AI power cost pledge, sources say,” Reuters, 13 July 2026)

The US: Trump channels crypto windfall into stocks and bonds as traditional holdings surge
On 13 July, Reuters reported that US President Donald Trump’s latest financial disclosures reveal that his money managers channeled significant proceeds from his family’s cryptocurrency ventures into traditional stocks and bonds, according to a Reuters analysis of filings with the US Office of Government Ethics. Trump reported receiving more than USD 1.4 billion last year from family-backed digital asset projects, including World Liberty Financial and a Trump meme coin. Following the influx of crypto revenue, the president's portfolio of traditional financial instruments increased at least fourfold, rising to a value between USD 703 million and USD 2.6 billion at the end of 2025, up from USD 225 million to USD 608 million the previous year. Digital asset experts noted that while Trump publicly champions making the US the "crypto capital of the world," his personal investment strategy favors safer harbours. The reallocation occurred even as retail investors suffered USD 2.3 billion in losses across Trump-backed crypto projects. The White House stated that Trump’s assets are held in discretionary accounts managed by independent institutions, while the Trump Organization defended the disclosure as evidence of a strong, conservative balance sheet with substantial liquidity.
(“Trump invested crypto gains in stocks and bonds, filings show,” Reuters, 13 July 2026)

India: TCS to hire nearly 8,900 AI engineers, eyes acquisitions
On 12 July, Reuters reported that Tata Consultancy Services (TCS) is building a team of up to 8,900 forward-deployed engineers. It is also seeking AI acquisitions, betting that AI will create new business rather than undermine outsourcing. The strategy comes amid investor concern that AI could disrupt India's USD 315 billion IT services industry. This includes reduced demand for engineering teams, shorter project timelines, and pricing pressure. Chief Executive Officer K Krithivasan said the forward-deployed engineer group would represent one to 1.5 per cent of the workforce, based on end-June headcount. He did not say whether roles would be filled externally or through reskilling. Krithivasan rejected concerns that AI would undermine outsourcing. He said deep knowledge of the customer environment is what makes it work, and that this is not about cost arbitrage. Chief Financial Officer Samir Seksaria said the company is also looking at acquisitions to enhance its strategic positioning. TCS's annualised AI revenue growth slowed to 13 per cent in the June quarter, from 28 per cent previously.
("India's Tata Consultancy Services plans up to 8,900 AI deployment engineers, seeks AI acquisitions," Reuters, 12 July 2026)

SK Hynix stock crashes 15 per cent as investors book profits after blockbuster Nasdaq debut
On 13 July, Reuters reported that SK Hynix shares fell more than 15 per cent in Seoul on Monday. This was the company's biggest one-day decline in nearly two decades, as investors cashed out of a sharp rally following its Nasdaq debut last week. The company's US-listed shares dropped 9.2 per cent to USD 152.50 in premarket trading on Monday. This followed a jump of over 12 per cent in its Nasdaq debut on Friday. Losses in SK Hynix and rival Samsung Electronics dragged South Korea's Kospi down 9 per cent, triggering a 20-minute trading halt. SK Hynix had raised over USD 26 billion last week by selling American Depositary Receipts (ADRs) priced at USD 149 each. The ADRs opened 14 per cent above the offer price at USD 170, before ending their debut session up 12.8 per cent. Morningstar director Lorraine Tan said monetisation remains uncertain despite growing AI adoption, and profitability for key players such as OpenAI appears under pressure. Chief Executive Officer Kwak Noh-jung dismissed concerns over aggressive capacity expansion. He said the memory industry is heading toward its most severe supply shortage in 2027.
 ("SK Hynix plunges after Nasdaq debut amid diminishing earnings optimism," Reuters, 13 July 2026)

Africa: Boko Haram, ISIS fighters uses AI chatbots for battlefield tactics
On 13 July, Indian Express reported that generative AI tools are increasingly aiding terrorist groups directly on the battlefield. This is despite efforts by AI companies to prevent misuse, according to a new research paper. The paper is based on nearly 60 interviews with 27 former members of Boko Haram in Nigeria, conducted over the past year by Cambridge researcher Antonia Juelich. Until recently, groups like the Islamic State and Al Qaeda mainly used AI for propaganda, translation, and recruitment. This has evolved as fighters increasingly seek tactical advantages on the ground, according to current and former counterterrorism officials. Terrorism expert Daniel Byman of Georgetown University said groups are combining different AI systems to bypass safety guardrails set by AI companies. He co-authored a related report released by the Center for Strategic and International Studies (CSIS). The research found Boko Haram used multiple AI platforms interchangeably. It said Islamic State Khorasan has urged followers to use AI to help avoid detection by authorities.
 ("How terrorist groups are using AI to gain an edge in battle," Indian Express, 13 July 2026)

11 July, Saturday
Apple sues OpenAI over alleged trade secrets theft to bolster hardware ambitions
On 11 July, Reuters reported that Apple filed a lawsuit against OpenAI and two former employees, Chang Liu and Tang Yew Tan, alleging a coordinated effort to steal trade secrets to accelerate OpenAI’s push into consumer hardware. The complaint alleges that Liu downloaded the confidential documents through an internal network bug and that, along with other confidential information, the company's hardware chief, Tan, shared information with suppliers and invited job hopefuls to bring proprietary Apple parts to interviews. The lawsuit also reflects a growing feud between the tech giants, as OpenAI has refuted the claims, saying it does not care about the trade secrets of its competitors. More than 400 of Apple's former staff members now work at OpenAI, and the company recently bought Jony Ive's hardware company, io Products, for USD 6.5 billion. Legal experts note that while hiring competitors' talent is legal in California, utilizing stolen data poses serious legal risks that could delay OpenAI’s ambitious hardware plans.
(“Apple sues OpenAI, two former employees for trade secrets theft,” Reuters, 11 July 2026)

10 July, Friday
Taiwan: Nanya unveils USD 6.2 billon expansion to meet surging AI memory demand
On 10 July, Reuters reported that Taiwan’s memory chipmaker Nanya Technology has announced plans for a massive preliminary capital expenditure of over TWD 200 billion (USD 6.2 billion) for 2027, roughly four times its projected 2026 spending. Driven by a massive structural shift in AI, the investment will fund a new manufacturing plant designed to alleviate an ongoing global memory chip supply shortage. This announcement follows an explosive second quarter for Nanya, which saw its revenue skyrocket 684 per cent year-over-year to TWD 82.55 billion and gross margins surge to 79.5 per cent. Serving high-profile tech giants like Nvidia, Google, and Qualcomm, Nanya anticipates the new plant's first phase will reach a monthly capacity of 30,000 wafers by 2028, eventually scaling to 45,000. Total production costs for the facility are estimated at TWD 480 billion, positioning Nanya to compete closely with South Korea’s rivals also aggressively expanding to meet AI-driven demand.
(“Taiwanese chipmaker Nanya plans $6 billion in spending in 2027, riding AI boom,” Reuters, 10 July)

South Korea: SK Hynix's USD 26.5 billion Nasdaq debut tests investor appetite for AI boom
On 10 July, Reuters reported that South Korea’s chipmaker SK Hynix is making its US trading debut on the Nasdaq following a massive USD 26.5 billion American Depositary Receipts (ADR) sale. Priced at USD 149 per ADR, the offering was seven times oversubscribed, marking the second-largest US share sale behind SpaceX’s recent IPO. As the global leader in high-bandwidth memory (HBM) chips, essential for Nvidia and AMD's AI processors, SK Hynix aims to utilize these funds to construct new factories and bridge the valuation gap with its US rival, Micron. This listing is a crucial litmus test for Wall Street's AI appetite. Arriving just as a recent 25 per cent pullback in the company's stock sparked broader questions about the sustainability of massive AI infrastructure spending. While global AI capital expenditure is projected to approach USD 1.5 trillion by 2027, the debut highlights the tension between AI enthusiasm and cyclical oversupply fears.
(“SK Hynix's marquee US debut to test AI appetite,” Reuters, 10 July 2026)

The EU: Meta risks fines as Brussels demands changes to social media engagement features
On 10 July, Reuters reported that the EU has charged Meta Platforms with breaching its landmark Digital Services Act (DSA), targeting Instagram and Facebook features designed to maximize engagement. EU tech regulators found that highly personalized recommendations, autoplay, and infinite scroll create excessive and addictive user behavior, contributing to the youth mental health crisis. The European Commission is demanding Meta alter these engagement-driven algorithms, disable autoplay and infinite scroll by default, and implement more rigid screen-time breaks. Regulators dismissed Meta's existing time management and parental controls as ineffective and easily bypassed.  Meta strongly disputes the preliminary findings, pointing to its rollouts of Teen Accounts with automated 15-minute screen time limits and parental overrides. However, the company must now modify its core platforms or face formal non-compliance decisions. Failure to resolve these issues could result in substantial financial penalties of up to 6 per cent of Meta's global annual turnover.
(“EU tells Instagram, Facebook to change addictive features or risk fines,” Reuters, 10 July 2026)

The UK: Paris designates major cloud providers as critical suppliers to finance sector
On 10 July, Reuters reported that the UK government has officially designated major cloud service providers including Microsoft, Google, Amazon, and Oracle, as "critical third-party suppliers" to its financial sector, effective 13 July. The move is part of a regulatory push to curb system risks from potential cyberattacks or broad tech failures, which is linked to the growing adoption of cloud services among banks and insurers. The new rules will see the tech giants directly targetted by the Bank of England, Prudential Regulation Authority (PRA) and Financial Conduct Authority (FCA), which will form a three-way joint watchdog. The companies will legally have to carry out periodic resilience testing, periodic self assessments and will have to notify customers promptly of major incidents to ensure long term operational resilience. While this strategy mirrors elements of the European Union's broader digital regulations, it marks a highly targeted effort by British authorities to safeguard the foundational stability of the UK's financial ecosystem.
(“UK to regulate cloud service providers Microsoft, Google and others to protect financial stability,” Reuters, 10 July 2026)

Biomni, an AI assistant, helps researchers turn biomedical data into hypotheses
On 10 July, South China Morning Post reported that Stanford University researchers from Stanford University introduced Biomni, an artificial intelligence assistant that is completely free and will assist scientists in analyzing biomedical data and creating hypotheses. Unlike other chatbots that respond to user queries, Biomni analyzes the information from scientific papers, genome databases, and laboratory tools for performing complex studies. The platform can analyze enormous amounts of biological data, detect patterns and generate areas of potential research. This significantly reduces the time required for conducting biomedical research. Biomni's developers believe that this platform will increase the pace of drug development, disease research, and precision medicine. Moreover, it makes sophisticated AI affordable for researchers who have limited funds.
("Meet Biomni: the free, powerful biomed AI agent turning data into hypotheses" South China Morning Post, 10 July 2026)

Token economy emerges as AI adoption accelerates in China 
On 10 July, South China Morning Post reported that according to experts from the Chinese tech industry, the formation of a new "token economy" is seen as a result of growing adoption of AI. Speaking at the AI conference, the experts claimed that AI tokens—tokens used by AI models to exchange information—become the main factor measuring the demand on computing and business value. Now companies use the amount of tokens consumed by their systems for pricing AI services, instead of the classic subscription to software. The growing consumption of large language models is expected to become the main economic indicator for AI investments, infrastructure development and competition in the rapidly growing AI market in China.
(“Token economy emerging as AI use soars in China, experts tell conference,” South China Morning Post, 10 July 2026)

Chinese AI firms invest in custom chips to reduce long-term costs
On 10 July, South China Morning Post reported that Chinese AI labs have stepped up efforts to develop their own AI chips even though they require heavy investment. Firms see these chips as being cost-effective in the long run and will also ensure better performance and reduce reliance on foreign semiconductor providers as export controls continue. The creation of such high-end chips requires heavy financial resources as well as technical expertise and several years of development process; however, top AI firms see chip autonomy as necessary for sustaining more sophisticated AI models and developing China's domestic semiconductor market.
(“Chinese AI labs pursue custom chips to lower costs despite heavy upfront investment risk,” South China Morning Post, 10 July 2026)

09 July, Thursday
Meta plans custom AI chip production, expands computing infrastructure
On 09 July, Reuters reported that according to an internal memo, Meta Platforms plans to begin manufacturing its custom-designed AI chip, code-named "Iris," in September. Developed alongside Broadcom and TSMC, the chip aims to improve Meta's social platforms, lower computing costs, and reduce dependence on dominant suppliers like Nvidia and AMD. The initiative supports Meta’s massive infrastructure scaling, which aims to double its computing capacity from seven gigawatts this year to 14 gigawatts next year. To power this rapid growth, Meta expects to spend up to USD 145 billion on AI infrastructure this year alone. To mitigate global hardware shortages and rising "chipflation," Meta has secured critical, multi-year supply agreements. These include deals with Samsung for memory chips, Sandisk for flash storage, and Sumitomo Electric for fiber-optic equipment, locking in necessary components for its rapidly expanding data centers.
(“Meta to put AI chip into production in September as it looks to double computing capacity, memo shows,” Reuters, 09 July 2026)

South Korea: SK Hynix US share sale draws strong investor demand, funds AI expansion
On 09 July , Reuters reported that according to a source familiar with the matter, demand for SK Hynix’s USD 28 billion US share sale was more than seven times oversubscribed, highlighting immense investor appetite for the AI supply chain leader. The offering, set to price on Thursday and begin trading on the Nasdaq on 10 July, will trail only SpaceX’s recent USD 85.7 billion IPO as the world's second-biggest share sale. The South Korea’s chipmaker plans to use the funds to build new factories and purchase equipment to meet surging global demand for its high-bandwidth memory (HBM) chips. Beyond financing infrastructure, the US listing is expected to help SK Hynix close its valuation gap with American rival Micron by gaining direct access to the world’s largest pool of investors. Driven by insatiable AI demand and a strong partnership with Nvidia, SK Hynix's stock is up 680 per cent over the past year. Massive earnings growth has outpaced these stock gains, lowering its 12-month forward price-to-earnings ratio to 5.5 times.
(“SK Hynix's US listing more than seven times oversubscribed, source says,” Reuters, 09 July 2026)

The UK: Palantir challenges police contract decision, trial set for 2027
On 09 July, Reuters reported that the US tech firm Palantir is legally challenging a decision by London Mayor Sadiq Khan's office to block a two-year, GBR 50 million (USD 67.06 million) contract with the Metropolitan Police. The contract aimed to integrate Palantir’s AI systems for automated administrative tasks and criminal evidence analysis. Palantir argues that the mayor's office unlawfully rejected the deal based on the company's "values and ethics," criticizing the move as "putting politics above public safety." Palantir has long faced Europe’s scrutiny over its US military and immigration software ties, alongside the political views of co-founder Peter Thiel. Conversely, the Mayor’s Office for Policing and Crime denies these allegations. In court filings, they maintained that the deal was blocked purely on procurement grounds because the Met Police failed to hold an open competition and only negotiated with a single supplier. London High Court Judge Adam Constable ruled that the trial will take place in January 2027.
(“Palantir says UK police contract wrongly blocked over perceived 'values',” Reuters, 09 July 2026)

France: Antitrust probe into Nvidia nears completion
On 09 July, Reuters reported that France’s competition authority announced that its antitrust investigation into AI chipmaker Nvidia Corp is nearing its end. The regulatory agency has been probing the company since complaints were made against it for anti-competitive activity in the semiconductor market. Umberto Berkani, the general rapporteur for the French authority, confirmed the timeline to reporters, noting that the probe is in its final stages. No specific details regarding potential findings, penalties, or exact conclusion dates were released.
(“French competition authority's Nvidia probe nearing end,” Reuters, 09 July 2026)

Taiwan: Central bank warns of AI bubble risks, holds interest rates steady
On 09 July, Reuters reported that during a parliamentary hearing, Taiwan’s central bank governor, Yang Chin-long, warned that while the economic expansion driving the AI boom is genuine, the risk of an AI bubble is equally real.  Yang acknowledged that AI has become a primary engine for Taiwan’s economy, anchored by giants like Taiwan Semiconductor Manufacturing Co (TSMC). However, he emphasized that the central bank must closely monitor the threat of over-expansion and speculative capital expenditures, which are being heavily financed by aggressive corporate borrowing and leverage within the technology sector. Despite these concerns, Taiwan's central bank board chose to keep interest rates steady at its June meeting. Yang defended the decision as appropriate, explaining that inflationary pressures from the tech sector were not yet severe enough to warrant a rate hike, especially given that Taiwan's traditional industries continue to lag behind the booming AI economy. 
(“Taiwan central bank chief warns of AI bubble risk,” Reuters, 09 July 2026)

The EU: Draft electrification plan targets reduced oil and gas dependence
On 09 July, Reuters reported that a draft European Commission proposal reveals that the European Union plans to introduce an extensive "electrification" strategy to rapidly shift its economy away from oil and gas. The plan, due to be officially published on 17 July, is a response to the energy crisis caused by Iran's involvement in the war, which has added a whopping EUR 50 billion (USD 57.11 billion) to the EU's import bill since late February when fossil fuel prices started to rise. To tackle this weakness, Brussels plans for 2040 to impose a minimum percentage target for electricity use in the economy as a whole. The idea will push the transition to EV's and domestic heat pumps forward, and take firm steps towards replacing industrial processes with electric furnaces. To address the high upfront costs of transition, the Commission will put in place mechanisms enabling member States to reduce the VAT rates on EVs, household batteries, and heat pumps. Also the EU will be removing fossil fuel subsidies later this year and implementing new public procurement regulations, which will require clean energy solutions in public buildings and public transport fleets.
(“EU drafts 'electrification' plan to curb oil and gas use, after Iran war disruption,” Reuters, 09 July 2026)

08 July, Wednesday
Crypto industry accelerates post-quantum transition as encryption risks near
On 08 July, Reuters reported that the cryptocurrency industry is drawing up plans to defend against quantum computing. This move follows recent technological advances that indicate standard encryption could be breached sooner than anticipated. Research from Google suggests quantum computers capable of cracking conventional cryptographic security could arrive by 2029, significantly compressing previous timelines. This poses an existential threat to the USD 2 trillion global digital asset market, which relies on decades-old elliptic-curve cryptography to secure transactions. Bitcoin is considered uniquely vulnerable due to its transparent transaction history, with research estimating that 35 per cent to 50 per cent of its circulating supply could be exposed to quantum attacks. While networks like Ethereum and Algorand are developing post-quantum roadmaps, experts warn that upgrading decentralized blockchains will be a multi-year, Y2K-style overhaul. The transition to larger, quantum-resistant digital signatures faces steep engineering hurdles, including increased storage requirements and higher transaction costs.
(“Crypto firms prepare defences as quantum threat to encryption draws nearer,” Reuters, 08 July 2026)

France: Paris reinforces EU publisher rights in Meta compensation dispute
On 08 July, Reuters reported that France's competition authority ordered Meta Platforms to resume talks and present a payment plan within 15 days to France’s media groups seeking a year of unpaid online publishing fees. The antitrust dispute, brought by France’s media associations DVP and APIG, centers on "neighbouring rights," an EU framework allowing publishers to seek compensation for digital use of their content. The associations alleged that Meta abused its dominant position by trying to dictate its own fee calculation methodology while withholding the information necessary to evaluate fair remuneration. A previous agreement between Meta and the publishers expired in 2024, leaving French media outlets without payments since 2025 despite their content remaining active on Meta's platforms. The antitrust authority opted not to issue a fine at this stage, noting it wanted to avoid setting a focal point that could lock the negotiations. While Meta stated it disagreed with the decision, the tech giant confirmed it would engage with the mandated process.
(“French antitrust watchdog orders Meta to resume talks with media groups over publishing fees,” Reuters, 08 July 2026)

China: Beijing intensifies AI security oversight with warning against Anthropic's Claude code
On 08 July, Reuters reported that a cybersecurity platform operated by China's Industry Ministry issued a warning regarding a serious security "backdoor" risk identified in Anthropic's AI coding assistant, Claude Code. The National Vulnerability Database (NVDB) stated that Claude Code versions 2.1.91 through 2.1.196 contain a built-in monitoring mechanism. This mechanism is capable of transmitting sensitive user information, including geographic locations and identity-related data, to remote servers without user consent, according to the database. It advised organizations and individual users to immediately review their systems to uninstall the impacted versions or upgrade to the latest secure release. It also urged tighter controls on external network access for development tools. The alert follows a recent decision by China’s e-commerce giant Alibaba to ban employees from using Claude Code after the software drew internal scrutiny. Anthropic did not immediately respond to requests for comment.
(“Anthropic says Alibaba illicitly extracted Claude AI model capabilities,” Reuters, 08 July 2026)

Russia: Moscow expands electronic warfare to disrupt Ukraine's Starlink-dependent drone operations
On 08 July, Reuters reported that Russia’s forces are deploying specialized electronic warfare systems utilizing civilian camouflage to counter Ukraine’s highly effective "mid-strike" drone campaign, as revealed by the Ukraine’s commanders. The low-cost, precise Ukraine’s drones rely heavily on Elon Musk's Starlink satellite internet network to communicate with remote pilots while striking logistics lines, fuel depots, and command centers dozens of kilometers behind the front lines. To disrupt this connectivity, Russia has begun deploying ground-based jamming systems, such as the Volna Kupol Garant, which can destabilize Starlink signals across an area of roughly 20 square kilometers.  Ukraine’s forces, including the 422nd Unmanned Systems Regiment, have prioritized hunting down and destroying these multi-trailer jamming installations to restore network functionality for their fleets. Concurrently, Russia’s forces have adapted their logistics by moving fuel and ammunition in small, decentralized convoys utilizing civilian transport, such as disguised milk trucks, commercial cars, and quad bikes, to avoid aerial detection.
(“Russia tries to jam Musk's Starlink systems to counter Ukrainian drones,” Reuters, 08 July 2026)

Nvidia strengthens CPU ambitions as Perplexity adopts Vera architecture
On 08 July, Reuters reported that AI startup Perplexity confirmed plans to deploy Nvidia’s new "Vera" central processing units (CPUs). The partnership is a significant win for the chip giant as it expands into a general-purpose processor market traditionally dominated by Intel and Advanced Micro Devices (AMD). Nvidia expects its Vera CPUs, designed as more generic computing units compared to its core AI graphics processors to generate USD 20 billion in sales by the end of the current fiscal year. The CPU line helps Nvidia diversify its revenue streams as prominent AI firms, including OpenAI and DeepSeek, move toward developing their own in-house custom AI chips. Unlike standard human infrastructure demands, autonomous AI "agents" operate continuously without breaks. Perplexity Vice President for Computer Enterprise and Infrastructure Nate Kupp stated that Nvidia's architecture handled agentic coding tasks approximately 1.5 times faster than traditional x86 CPUs, describing the technology as a precise fit for its core workloads. While Perplexity declined to specify the volume of its order, Nvidia has previously stated that OpenAI, Anthropic, and Oracle also intend to adopt the new hardware.
(“Perplexity says it plans to use Nvidia's new CPU,” Reuters, 08 July 2026)

Think tank warns of gap in AI sector on safety front
On 08 July, Japan Today reported that an influential AI think tank has expressed serious concern that the global AI sector has not been able to keep up with the necessary safety aspects as the models become increasingly potent and get rolled out swiftly. The think tank has pointed out that several firms have continued to emphasise innovation and competitiveness over thorough tests, risk evaluations and independent reviews. The think tank has emphasised the importance of greater collaboration across borders and common safety guidelines to minimise the dangers of advanced AI technologies.
(“Global AI industry falls short on safety, think tank warns,” Japan Today, 08 July 2026)

MIT develops AI tool for non-expert military programmers
On 07 July, MIT News reported that scientists from the Massachusetts Institute of Technology have created an AI tool that would allow inexperienced programmers to develop AI software for military purposes without any sophisticated programming knowledge. The technology consists of combining natural language commands with automation software that will help users to create AI software faster and better. Scientists see potential in the technology as it may increase the speed of developing defence-related technologies while making the process of creating AI software simpler. The project is another example of how generative AI may help to automate complicated tasks and let more people create their special AI software.
(“How novice coders can develop AI programs for military applications,” MIT News, 07 July 2026)

07 July, Tuesday
China: DeepSeek advances in-house AI chip development amid export controls
On 07 July, Reuters reported that China’s AI champion DeepSeek is developing its own custom chip specifically designed for AI inference, the stage where trained models generate user responses. The initiative aims to reduce the startup's reliance on hardware from Nvidia and Huawei, crucial for navigating tightening US export controls that restrict China's access to advanced foundries and critical high-bandwidth memory. The project began about a year ago and remains in its early stages. DeepSeek has quietly ramped up hiring for chip-design engineers and opened discussions with external foundry and memory partners. By building in-house hardware, DeepSeek is following global competitors like OpenAI and Anthropic seeking greater infrastructure control. This major strategic pivot coincides with DeepSeek’s first-ever external funding round, in which it plans to raise USD 7 billion, valuing the company at up to USD 59 billion.
(“China's DeepSeek developing its own AI chip, sources say,” Reuters, 07 July 2026)

China: Beijing considers curbs on overseas access to advanced AI models
On 07 July, Reuters reported that China’s authorities have held meetings with top technology firms, including Alibaba, ByteDance, and startup Z.ai, to discuss restricting overseas access to the country’s most advanced AI models, three sources familiar with the matter said. The discussions, led by the Ministry of Commerce, mark a major step by Beijing to treat cutting-edge artificial intelligence as a critical national asset. Officials warned that leaks or theft of proprietary AI could become punishable offenses under China’s stringent national security law. New curbs on funding for domestic AI startups were also raised. The potential restrictions follow global success for low-cost Chinese models, like DeepSeek's R1 and Z.ai's GLM-5.2. They mirror similar actions by the US, which recently restricted access to Anthropic's advanced models over national security concerns, highlighting an intensifying global tech divide.
(“Beijing is looking at curbing overseas access to China's top AI models, sources say,” Reuters, 07 July 2026)

China: Beijing records surge in new unicorn start-ups driven by AI
On 06 July, the South China Morning Post reported that China witnessed a surge in its fastest rate of new unicorn startups in five years. A unicorn refers to a privately owned company valued at more than USD 1 billion. In the current growth, there have been a total of 67 unicorns established. Over 50 per cent of the newly formed ventures belong to the realm of artificial intelligence and robotics. This growth is indicative of China’s huge investments in technological advancements. The establishment of such technologically oriented ventures has taken place despite geopolitical challenges and export restrictions.
(“China records most new unicorn start-ups in 5 years as AI and robotics boom,” SCMP, 06 July 2026)

China: Hong Kong makes AI a central plank for economic growth
On 07 July, the South China Morning Post reported that Hong Kong's Chief Executive, John Lee, stated that artificial intelligence would be an essential element of the city's economic development. Speaking during an economic forum, Lee explained that the upcoming five-year plan would place high priority on AI-based development. The government wants to capitalise on the city's advantages to facilitate cooperation between Chinese technology companies and the international community. This would enable Hong Kong to attract the best tech specialists and considerable investments. The city expects AI development to modernise its economy by building special centres.
(“AI 'central plank' in Hong Kong's economic development, John Lee says,” SCMP, 07 July 2026)

The US: Meta faces USD 1.4 trillion youth safety lawsuit
On 06 July, Reuters reported that Meta Platforms disclosed in a court filing that four US states are seeking USD 1.4 trillion in penalties, accusing the company of intentionally designing Facebook and Instagram to addict young users. The previously undisclosed figure, nearly matching Meta’s USD 1.5 trillion market capitalization comes ahead of an August trial in Oakland, California. The lawsuit, led by California, Colorado, Kentucky, and New Jersey, alleges Meta misled the public about platform safety and fueled a youth mental health crisis. Meta rejected the damages as unsupported by evidence, stating a sanction of that size has no analog in consumer protection history. The states calculated the penalties by multiplying the estimated number of affected teenage users by statutory fine amounts under state laws. The trial before US District Judge Yvonne Gonzalez Rogers is part of a broader legal wave, with Meta facing thousands of similar lawsuits nationwide.
(“Meta says US states are seeking $1.4 trillion in penalties in August youth safety trial,” Reuters, 06 July 2026)

The US: CISA adopts Anthropic AI to strengthen government cybersecurity
On 07 July, Reuters reported that the US Cybersecurity and Infrastructure Security Agency (CISA) is using Anthropic’s advanced AI model, Mythos, to audit government software, three people familiar with the matter said on Monday. The agency's Attack Surface Evaluation team is employing the specialized model to scan federal code repositories for bugs that could be exploited by foreign spies or cybercriminals. Two sources noted that the audits have already uncovered a substantial number of vulnerabilities. The adoption highlights growing government reliance on Anthropic's tools despite a rocky relationship with the White House. Anthropic faced a brief Pentagon blacklist in February over safety protocol disputes, but national security agencies have increasingly embraced Mythos due to its powerful ability to detect cyber flaws. The NSA has reportedly been testing the model in classified environments, even amid recent White House export restrictions that briefly triggered a global shutdown of its public counterpart, Fable.
(“US cyber agency is using Anthropic's Mythos to audit government code, sources say,” Reuters, 06 July 2026)

Nigeria: Abuja investigates global tech firms over news content and AI practices
On 07 July, Reuters reported that Nigeria’s President Bola Tinubu has directed the country’s antitrust regulator to investigate global tech giants over alleged anti-competitive practices and the unauthorized use of local news content, the Federal Competition and Consumer Protection Commission (FCCPC) announced on 06 July. The inquiry stems from a formal complaint by the Nigerian Press Organisation, which represents newspaper owners, broadcasters, journalists, and online publishers. The FCCPC stated it will investigate Alphabet, Meta, X, and various generative AI platforms operating in the country. The probe will scrutinize allegations of market dominance, the unauthorized extraction of copyrighted journalistic material for commercial gain, and the scraping of local news content to train generative AI models. The investigation follows a growing global trend of regulators pushing tech firms to compensate publishers. In Africa, South Africa recently secured a ZAR 688 million (USD 42 million) media support package from Google following a similar market inquiry.
(“Nigeria to investigate tech firms over news content use,” Reuters, 07 July 2026)

The UK: Bank of England warns of growing financial stability risks posed by AI
On 07 July, Reuters reported that The Bank of England (BoE) warned that rapid advancements in AI pose a growing threat to global financial stability, citing over-leveraged tech investments and heightened cyber risks. In its half-yearly Financial Stability Report, the central bank noted that heavily concentrated, momentum-driven market positions in AI-related stocks could amplify volatility if future earnings fail to meet expectations. Furthermore, AI-related companies are borrowing heavily to fund infrastructure, creating hidden debt risks due to a lack of market transparency. Operationally, the BoE highlighted systemic vulnerabilities introduced by autonomous "agentic" AI systems that act with minimal human intervention. Deputy Governor Sarah Breeden recently emphasized the need for bespoke regulations to manage these models. Additionally, the BoE noted that keeping pace with AI-driven cyber threats will require financial institutions to implement more frequent software updates, increasing the risk of widespread operational disruptions.
(“Bank of England sees growing risks to financial stability from AI,” Reuters, 07 July 2026)

Investors may shift focus from AI chipmakers to big tech firms, reports Reuters
On 06 July, Reuters reported that Morgan Stanley expects that attention will eventually shift from semiconductor shares to AI hyperscalers such as Alphabet, Amazon, and Meta. The firm is confident that the benefits from the growth of the AI infrastructure are already fully realised by chipmakers, but now hyperscalers will be able to generate more interest because they concentrate on maximising profits from their enormous investments in AI. Additionally, Morgan Stanley anticipates more discipline in  expenditures related to AI from companies.
(Shashwat Chauhan, “AI investors may pivot to hyperscalers from chipmakers, Morgan Stanley says,” Reuters, 06 July 2026)

06 July, Monday
South Korea: SK Hynix launches USD 28 billion Nasdaq listing amid AI chip boom
On 06 July, Reuters reported that South Korea’s chipmaker SK Hynix launched a massive US share sale to raise KRW 43 trillion (USD 28.07 billion) on the Nasdaq. This move aims to capitalize on global AI demand. The company will reportedly sell 17.79 million new shares through American Depositary Receipts (ADRs), with 10 ADRs representing one common share. Major investors, including Baillie Gifford Overseas, Coatue Management, and Situational Awareness Partners, have already indicated interest in buying up to a combined USD 7 billion of the offering.The final listing price is expected to be held on 09 July ahead of 10 July’s trading. The move comes amid a broader memory super-cycle driven by AI, alongside a newly announced USD 576 billion South Korea’s government chip investment program. Despite shares dipping 3.4 per cent amid recent market volatility, SK Hynix stock remains up approximately 260 per cent this year.
(“South Korea's SK Hynix launches $28 billion US listing to ride global AI wave,” Reuters, 06 july 2026)

Samsung set for record Q2 profit on AI-driven memory demand
On 06 July, Reuters reported that Samsung Electronics is projected to report an 18-fold jump in second-quarter operating profit to KRW 86 trillion (USD 56.35 billion), marking its third consecutive record-breaking quarter. The surge is driven by a prolonged memory chip shortage and skyrocketing prices fueled by the global AI boom, particularly the expansion of "agentic AI" workloads. DRAM and NAND average selling prices jumped 44 per cent and 53 per cent respectively in Q2, boosting Samsung's stock by 158 per cent this year. However, analysts note potential earnings headwinds from massive employee bonus provisions following a recent wage deal. Long-term risks include potential slowdowns in cloud provider AI spending, which could impact Samsung's plans to invest KRW 3,200 trillion (USD 2.07 trillion) in domestic chip capacity through 2040. Additionally, rising component costs are currently squeezing profit margins in Samsung's mobile division.
(“Samsung likely to post 18-fold jump in profit on surging AI demand for memory,” Reuters, 06 July 2026)

UN urges global AI governance as adoption outpaces regulation
On 06 July, Reuters reported that the UN Secretary-General Antonio Guterres warned that AI is outrunning global regulation, further urging harmonized frameworks to mitigate risks, particularly to children. Speaking at the first government-level Global Dialogue on AI Governance in Geneva, Guterres noted that AI reached one billion users in just two years, far outpacing the internet's early adoption.  Guterres called for a mandatory "AI Child Safety Pledge," requiring companies to prove system safety before minor accessibility, demanding a complete halt and connection to human support if a child exhibits distress.  Additionally, an independent expert report highlighted a severe geopolitical AI divide: the US and China hold 75 per cent and 15 per cent of top AI supercomputing power respectively, threatening to lock developing nations out of the tech's benefits. African and European leaders reinforced calls to expand infrastructure and prevent AI from becoming an instrument of digital tyranny. 
(“UN's Guterres warns AI outpacing oversight, urges global rules to protect children,” Reuters, 06 July 2026)

South Korea: Seoul fast-tracks USD 576 billion chip and AI investment plan
On 06 July, Reuters reported that South Korea’s President Lee Jae Myung ordered officials to fast-track execution of the country's newly announced USD 576 billion chip and AI investment plan, warning that administrative delays could damage South Korea's competitiveness. Emphasizing that "only speed matters" in the global advanced industry race, Lee directed agencies to run sequential regulatory procedures in parallel and streamline environmental reviews. The mega-project includes separate KRW 400 trillion (USD 260 billion) semiconductor manufacturing sites by Samsung Electronics and SK Hynix in the southwestern Gwangju region, alongside an KRW 81 trillion packaging cluster in Chungcheong. Lee specifically targeted infrastructure bottlenecks, instructing officials to pre-emptively secure land, water, and baseload power supplies to address corporate grid concerns despite expanding renewable energy capacity.
(“South Korea's Lee urges speed in launching mega chip projects,” Reuters, 06 July 2026)

Denmark and Belgium: Copenhagen backs Brussels in EU court battle over publishers’ rights
On 06 July, Reuters reported that Denmark’s government announced that it has filed a written intervention in the European Court of Justice (ECJ) supporting Belgium in a high-stakes legal battle against major tech companies over press publishers' rights. The case stems from a 2023 lawsuit filed by Google, Meta, Spotify, Sony, and Streamz, who argue that Belgium's implementation of Article 15 of the EU's Digital Single Market Directive undermines EU law. Denmark will join the oral hearings on 06-07 July to argue that tech giants must remain legally and financially accountable for using newspaper articles and media content on their platforms. Danish Culture Minister Zenia Stampe warned that a ruling in favor of the tech companies could dilute publishers' protections, ultimately harming local media and democracy. The ministry noted Denmark is also actively participating in a separate landmark European copyright case regarding Google's training of artificial intelligence on press materials.
(“Denmark intervenes in EU court case over publishers' rights,” Reuters, 06 July 2026)

04 July, Saturday
Australia: NSW backs OpenAI investment amid growing environmental concerns over AI data centres
On 04 July, the Guardian reported on the New South Wales government's internal emails revealing that officials softened their public response to OpenAI's decision to establish its first Australian office in Sydney. The staff compared the company's AI ambitions to the fictional "Skynet" from the Terminator films. Although references to being "absolutely thrilled" were removed from the official statement, the documents show the government actively encouraged OpenAI's investment by promoting Sydney as Australia's leading AI and technology hub. The revelations come as concerns grow over the rapid expansion of AI data centres. NSW environmental modelling warned that simultaneous use of diesel backup generators during a widespread power outage could produce pollution levels several times higher than emissions from the state's electricity generation and motor vehicles. While industry representatives described this as an unlikely worst-case scenario, critics argue the increasing number of data centres could strain the power grid and pose environmental and public health risks.
(“NSW government ‘absolutely thrilled’ to welcome OpenAI ... until someone mentioned the Terminator films,” The Guardian, 04 July 2026)

China: ByteDance discovers AI scaling law that could sustain autonomous agent development
On 04 July, South China Morning Post reported that researchers at TikTok’s parent ByteDance discovered a mathematical scaling law governing how fast AI agents improve by performing real-world tasks, potentially sustaining the AI boom as traditional training data sources face depletion. The findings, published by ByteDance’s Seed AI team, reveal autonomous AI agents double their learning speed every three months through extended interactions with real-world environments. Utilizing a new benchmarking suite called EdgeBench, researchers logged 38,000 hours of environment interaction to evaluate five frontier models, including Anthropic's Claude Opus 4.8 and OpenAI's GPT-5.5. The long-duration tasks forced the autonomous software to run continuously for 12 to 72 hours. The data demonstrated that post-deployment learning follows a highly predictable log-sigmoid mathematical curve.
(“China’s ByteDance discovers new scaling law that could sustain AI boom,” South China Morning Post, 04 July 2026)

The Netherlands: The Hague invests in unified drone software platform to enhance battlefield interoperability
On 03 July, Reuters reported that the Netherlands military launched a three-year partnership worth with a defense technology company Intelic worth tens of millions aiming at developing a unified drone software platform. The initiative aims at establishing a common command-and-control system allowing unmanned hardware from different manufacturers to seamlessly operate together on the battlefield. Junior Defense Minister Derk Boswijk noted that lessons from the war in Ukraine underscore the critical importance of software over hardware in modern conflict. "Making different drone systems work together makes the fight easier," Boswijk said. The platform will expand on Intelic’s NEXUS software, which has seen active battlefield deployment in Ukraine since last year. Intelic CEO Maurits Korthals Altes highlighted the urgency of the project, noting that Europe's rapid expansion to over 700 drone manufacturers makes software interoperability the primary challenge for modern defense organizations (“Dutch military invests millions in drone software platform,” Reuters, 03 July 2026)

03 July, Friday
Argentina: Buenos Aires proposes legal framework for AI-run companies; seeks to position itself as a global AI investment hub
On 03 July, Reuters reported that Argentina’s President Javier Milei introduced a congressional bill aiming to create a new category of "automated companies" run by AI. Milei pitched the "non-human corporations" as an innovative way to attract foreign investment and establish Argentina as a global AI hub. However, legal experts and the bill’s co-authors clarify that human involvement remains legally mandatory. Under the proposed reform, these companies must appoint a human administrator to oversee operations, and the business will remain fully liable for any damages caused by AI systems. Moreover, the bill also allows for the creation of decentralised autonomous organisations (DAOs) on the blockchain, though it introduces a contentious security requirement that token users be registered. While Silicon Valley entrepreneurs note that light-touch regulation could appeal to investors facing stricter rules in the US and Europe, tech experts emphasise that infrastructure, energy costs, and talent retention will ultimately determine Argentina’s success.
(“Argentina's plan for AI-run companies can't avoid humans,” Reuters, 03 July 2026)

The US: Trump administration denies talks on government stake in Anthropic amid growing debate over public ownership of AI firms
On 03 July, Reuters reported that the Trump administration and AI firm Anthropic have not engaged in discussions regarding a potential government equity stake in the company. The disclosure follows a Financial Times report revealing that competitor OpenAI has discussed giving the US government a five per cent stake. Leading AI firms face tightening oversight in Washington over national security concerns and public accountability according to Commerce Department's recent temporary export controls on Anthropic's advanced models. President Donald Trump previously noted he was exploring options to grant the public a financial stake in major AI ventures. However, Senator Bernie Sanders proposed a steeper 50 per cent tax-driven ownership mechanism. Despite OpenAI's separate proposal to include other developers in a public wealth fund model, Anthropic, the White House, and the Commerce Department have declined to comment on active negotiations.
(“Trump administration and Anthropic have not discussed the government taking a stake in it, source says,” Reuters, 03 July 2026)

Meta says AI agent development lags expectations despite increased spending
On 03 July, Reuters reported that Meta CEO Mark Zuckerberg admitted that the company's AI agent technology has progressed more slowly than anticipated. This trend follows a sweeping corporate restructuring. Zuckerberg reportedly noted that structural bets including laying off 10 per cent of Meta's global workforce and reassigning 7,000 employees to AI-focused teams “haven’t come to fruition yet." However, he expects significant benefits within the next three to six months. The social media giant, projected to spend up to USD 145 billion on AI infrastructure this year, is attempting to moderate recent organisational changes amidst employee pushback over morale and timing miscalculations. Additionally, Chief Technology Officer Andrew Bosworth addressed a recent security review regarding Meta's controversial, recently paused employee mouse-tracking software. Bosworth assured staff that no sensitive corporate or personal employee data was used in AI model training, adding that any future resumption of the data-collection program will switch to a strictly voluntary, opt-in basis.
(“Meta's Zuckerberg says AI agent tech progressing slower than expected,” Reuters, 03 July 2026)

China: Alibaba bans Anthropic's Claude Code over security concerns, deepening AI rivalry with the US
On 03 July, Reuters reported that China’s tech giant Alibaba has banned its employees from using Anthropic’s Claude Code assistant, classifying it as "high-risk software" effective 10 July. The directive reportedly instructs developers to immediately transition to Qoder, Alibaba's proprietary in-house AI coding platform.  The workplace ban intensifies an escalating geopolitical dispute after independent researchers revealed that Claude Code contained hidden, obfuscated logic. The software silently inspected local user environments specifically looking for China’s time zones, proxies, and major corporate networks like Alibaba and ByteDance. Moreover, it also used steganography to embed hidden tracking markers into prompt data sent back to Anthropic servers. Anthropic employees publicly defended the feature as an authorized "experiment" to combat illicit account reselling and prevent China’s firms from executing industrial-scale "model distillation" attacks to scrape its proprietary capabilities. However, Alibaba internally flagged the tracking mechanisms as a major enterprise security and data privacy backdoor.
(“Alibaba to ban employees from using Anthropic's coding tool, source says,” Reuters, 03 July 2026)

02 July, Thursday
China: Beijing AI drug-design deals swell despite mounting US scrutiny, notes HSBC
On 02 July, South China Morning Post reported that cross-border dealmaking by China’s AI-driven drug-design firms is expanding rapidly on the global stage. This expansion is driven by multinational pharmaceutical corporations seeking innovative assets before major patents expire. According to research from HSBC, out-licensing deals struck by Chinese biotech companies with global pharmaceutical firms reached USD 75 billion in the first five months of 2026, a surge from near zero before 2020. In total, these firms secured 169 licensing agreements worth a combined USD 93 billion during this period, marking an 87 percent year-on-year increase.  This commercial expansion continues despite tightening regulatory pressures from Washington, including a recently proposed US national security review targeting biotechnology investments in Chinese entities. Highlighting this trend, Hong Kong-listed Metis TechBio signed an exclusive cross-border agreement this week with Deerfield Management-backed Boulevard Bio. The deal, valued at up to USD 1.6 billion in milestone payments and a USD 20 million upfront fee, grants the US firm global commercialization rights to MTS-128, an AI-optimized next-generation cancer treatment candidate. 
(“China AI drug-design deals swell as US scrutiny mounts,” South China Morning Post, 02 July 2026)

China: New inexpensive AI model of China GLM-5.2 bridged the capability gap with OpenAI and Anthropic
On 02 July, Reuters reported that China-based startup Z.ai (Zhipu AI) launched GLM-5.2, an open-weight AI model that is rapidly closing the capability gap with the US tech giants, including OpenAI and Anthropic. Operating at roughly a sixth of the cost of premium Western rivals, GLM-5.2 is generating significant Silicon Valley buzz for its advanced coding and autonomous agent performance. The model currently ranks fifth on Artificial Analysis' global LLM intelligence leaderboard and has surged past Anthropic on developer platforms like OpenRouter. This "mini DeepSeek moment" is accelerating open-source adoption, particularly among tech startups eager to bypass the unpredictable costs of closed-source US models. However, widespread adoption among Western corporations remains restricted due to lingering data security and compliance concerns in heavily regulated industries like banking.
(“A new, inexpensive Chinese AI model is catching up with Anthropic, OpenAI on their home turf,” Reuters, 02 July 2026)

The US: OpenAI proposes five per cent government stake to Trump Administration, Financial Times reports
On 02 July, Reuters reported that, according to a Financial Times report, OpenAI has discussed giving the US government a 5 percent equity stake. The proposal comes amid mounting political scrutiny in Washington over the possibility of widespread economic upheaval and massive job losses due to the rapid AI boom. In addition, CEO Sam Altman has reportedly discussed the arrangement with President Donald Trump and key cabinet officials. Modeled on the Alaska Permanent Fund, the proposed structure would pool equity from leading US AI firms into a vehicle that could eventually pay public dividends. The strategy aims to manage regulatory uncertainty ahead of planned initial public offerings. It follows recent government interventions, including a temporary halt on OpenAI's GPT-5.6 rollout. Analysts warn that while a pre-IPO government stake could appease US regulators, it may prompt foreign jurisdictions to demand similar data sovereignty compromises as a condition for market access.
(“OpenAI proposes handing Trump administration 5% stake, FT reports,” Reuters, 02 July 2026)

09 June, Tuesday
The US: OpenAI files for confidential IPO after filing done by rival Anthropic
On 09 June, Reuters reported that OpenAI filed for a confidential Intial Public Offering (IPO), after rival Anthropic had filed theirs. This is seen as a push toward enhancing the exposure to gather investors in the stock market. Reuters reported, OpenAI did not disclose the size or terms of the offering, and said a timeline has ‌not yet been determined. The IPO involves OpenAI renegotiating with its initial investor Microsoft regarding their partnership. 
("OpenAI files for US IPO after Anthropic as AI giants head to public markets,” Reuters, 09 June 2026)
https://www.reuters.com/technology/openai-files-us-ipo-after-anthropic-ai-giants-head-public-markets-2026-06-08/) 

The US and China: Pentagon labels certain Chinese companies as ‘military companies’; blacklists Alibaba and others
On 09 June, South China Morning Post reported that the Pentagon has signalled it was adding Chinese companies like Alibaba, BYD, Baida to the blacklist. A Federal Registrar notice scheduled for Wednesday, the US Department of Defense listed a range of broad range of Chinese firms as Chinese military companies. These include electric vehicle makers, artificial intelligence companies, battery manufacturers, biotech firms and solar suppliers. The designation can complicate companies’ access to US capital markets and government business. 
(“US adds Alibaba, BYD and other Chinese tech champions to military company blacklist,” South China Morning Post, 09 June 2026)
https://sc.mp/gs5lr

The US: Trump administration denies retaliation in Anthropic AI blacklisting lawsuit
On 09 June, Reuters reported that the Trump administration has unlawfully denied retaliation against the AI firm Anthropic in a court filing. Acknowledging that federal agencies restricted the use of its products after the company resisted Pentagon demands regarding military applications of its Claude chatbot. In March, Anthropic had sued the administration, alleging that President Trump and Defense Secretary Pete Hegseth blacklisted the company in violation of its free speech and due process rights. According to the filing, Pentagon imposed a supply-chain risk designation as Anthropic refused to remove safeguards preventing its AI systems from being used for autonomous weapons and domestic surveillance. In return, the Department of Justice argued that the lawsuit should be dismissed on procedural grounds, claiming the designation is not subject to judicial review.
(“Trump administration denies unlawful retaliation in Anthropic AI blacklisting,” Reuters, 09 June 2026)
https://www.reuters.com/legal/litigation/trump-administration-denies-unlawful-retaliation-anthropic-ai-blacklisting-2026-06-09/  

Apple deepens AI push with new Siri and privacy-focused software upgrades
On 09 June, Reuters reported that, Apple has unveiled a major overhaul of its software ecosystem at its Worldwide Developers Conference (WWDC) 2026. This is spearheaded by the launch of Siri AI, an advanced AI-powered personal assistant system built on Apple Intelligence. More natural and human-like interactions are now possible with the new Siri, as it can understand personal context, analyse on-screen content, interpret images, search the web, and carry out tasks across apps. Apple also introduced Apple Intelligence to core applications including Safari, Messages, Mail, Calendar and Home, with a focus on privacy by processing at the edge. Plus, the company also added improved child-security features such as more robust parental controls, web access filtering and approval checks for web access. Apple has also enhanced its own image generating and editing capabilities in Image Playground and photo editing with AI-powered tools for creating and editing images.
(“From Siri AI to child safety tools: Key takeaways from Apple's WWDC,” Reuters, 09 June 2026)
https://www.reuters.com/business/siri-ai-child-safety-tools-key-takeaways-apples-wwdc-2026-06-08/