The US-Saudi Arabia Nuclear Cooperation Agreement I AI's Rogue Agent Breach

Photo Source:
NIAS Course on Global Politics
National Institute of Advanced Studies (NIAS)
Indian Institute of Science Campus, Bangalore
For any further information or to subscribe to GP alerts send an email to subachandran@nias.res.in

CWA # 2212, 26 July 2026

The World This Week

The US-Saudi Arabia Nuclear Cooperation Agreement I AI's Rogue Agent Breach

NIAS Global Politics Team

TWTW Note
The US-Saudi Arabia Nuclear Cooperation Agreement:
Questions over the agreement's terms, the Abraham Accords condition, and Non-proliferation concerns

Madhura Meenakshi Tanikella

What happened?
On 22 July, US Secretary of Energy Chris Wright and Saudi Arabia’s Minister of Energy His Royal Highness Prince Abdulaziz bin Salman signed a “peaceful nuclear cooperation agreement” known as the “123 Agreement,” alongside an accompanying bilateral safeguards agreement.

On 23 July, US President Trump stated that the deal with Saudi Arabia is a civil nuclear deal, does not include enrichment, and hinges on Saudi Arabia joining the Abraham Accords. Subsequently, Israel’s Prime Minister’s Office released a statement, saying that Saudi Arabia joining the Abraham Accords would be an “historic leap” without mentioning the nuclear deal.

Democratic Party members in the US have expressed concern over the deal, stating, “This agreement will set off a nuclear race in the region, further disincentivizing Iran from limiting its own program,” and that the US Congress would only approve the deal if it had “the same gold-standard safeguards as our deal with the UAE.”

What is the background?
1. A note on the “123 Agreement”
The "123 Agreement" refers to Section 123 of the US Atomic Energy Act of 1954, which requires Washington to conclude a peaceful nuclear cooperation agreement before any transfer of significant nuclear material, equipment, or technology to another country, ensuring compliance with global non-proliferation standards. The US currently has 26 agreements with different countries and international organizations. In the Middle East, the UAE and now Saudi Arabia are a part of this agreement. The US has used these agreements to strengthen strategic partnerships, promote a degree of nuclear non-proliferation, and establish dominance in the nuclear export trade.

2. The US-Saudi Arabia nuclear dialogue
Ten years ago, Riyadh and Washington began discussing a civil nuclear program and a potential 123 Agreement. Negotiations date back to the first Trump administration and continued through Joe Biden’s presidency, when a nuclear deal was linked to Saudi-Israel normalization. In 2025, talks were revived; during the November visit to the US by Crown Prince Mohammed bin Salman, the two sides signed a Joint Declaration on the Completion of Negotiations on Civil Nuclear Cooperation, laying the groundwork for the formal 123 Agreement.

3. Saudi Arabia’s long-standing nuclear ambitions
In 2019, Prince Abdulaziz said that the Kingdom aims to go "from producing uranium, to enriching uranium, using the uranium, even acquiring the technologies." In 2023, this was reiterated by him conveying the country’s intent to produce low-enriched uranium (LEU) domestically. The nuclear power is expected to free up oil for export and diversify the Kingdom's energy mix under Vision 2030. Externally, the rivalry with Iran has been a key factor. In 2018, Prince Mohamed bin Salman said that Saudi Arabia would follow Iran if it were to develop a nuclear bomb, “without a doubt.” Additionally, Saudi Arabia is reported to have helped Pakistan's nuclear weapons program in the 1970s and is speculated to have an understanding with Islamabad under the 2025 Strategic Mutual Defence Agreement that extends a "nuclear umbrella."

4. The Abraham Accords as a conditional factor
The Abraham Accords, the set of US-brokered agreements in 2020 between Israel and the UAE and Bahrain, later signed by Morocco, Sudan, and Kazakhstan, aimed to normalize diplomatic relations among the Arab countries and Israel. Saudi Arabia has long maintained that any normalization with Israel would require the establishment of a Palestinian state. Additionally, during the Israel-Gaza war in 2024, Mohammed bin Salman stated, “The Kingdom reiterates its condemnation and absolute refusal of the collective genocide committed by Israel against the brotherly Palestinian people,” while continuing to take a stand against Israel’s actions in the region as the US-Iran conflict persists.

5. Apprehensions for non-proliferation
The agreement has been criticized by non-proliferation experts for deviating from the "gold standard" set by the 2009 UAE 123 Agreement, in which the UAE voluntarily consented to forgo plutonium reprocessing and uranium enrichment. Their main argument is the current 123 agreement, which permits an enrichment option and exempts Riyadh from joining the IAEA Additional Protocol. According to experts in the field from Carnegie and Stimson, the agreement creates new, risky precedents at a time when nuclear weapons are expanding globally, and its provisions are "tantamount to the United States giving up on non-proliferation." These concerns come in the backdrop of increasing tensions in the Middle East.

What does this mean?
First, the Abraham Accords condition adds uncertainty. Saudi Arabia has not responded to Trump's statement, leaving it unclear whether this constitutes a binding part of the agreement, particularly given Riyadh's consistent linkage of Palestinian statehood to any normalization with Israel.

Second, should the reported non-proliferation standards prove to be true, this would reflect a shift in US policy, setting a precedent that could weaken established non-proliferation norms and also allow other countries like South Korea to advance similar interests.
Third, the deal is likely to shape the broader Middle East calculus, given that Saudi Arabia and the UAE are the only countries party to a 123 Agreement in the region, further feeding Iran's and Israel's strategic calculations and threat perceptions.

Madhura Meenakshi Tanikella is an undergraduate student from the Department of Political Science, School of Liberal Arts, Alliance University, Bengaluru. She is currently an intern at NIAS.

TWTW Note
AI's Rogue Agent Breach:
Recurring Patterns, Gaps in the AI safety rules and the AI rivalry

Priyanka Makhijani

What happened?
On 21 July, OpenAI disclosed that one of its AI agents went rogue during a security test, breaking out of containment and hacking into AI startup Hugging Face’s systems to complete its test. OpenAI called it an unprecedented cyber incident and said it’s strengthening its safeguard.

On 22 July, Hugging Face revealed that it turned to Zhipu AI's open-source model, GLM-5.2, to help analyze the hack. It did this after several leading US AI models refused the task because they could not tell whether they were dealing with a defender or the attacker.

What is the background?
1. Recurring breach patterns
In April 2026, Anthropic's next-generation model, Claude Mythos, was breached. Unauthorized users, including someone with contractor access, gained entry to the restricted model through a third-party vendor environment days after its launch. Mythos's unprecedented ability to autonomously find and exploit vulnerabilities raised alarm. In November 2025, Anthropic disclosed that the Chinese state-sponsored hackers manipulated its Claude models into autonomously executing 80 to 90 percent of an espionage campaign against 30 plus organizations. Together, these episodes show agentic AI risk recurring across multiple companies and models.

2. Goals over guardrails by the AI agents
Modern AI agents are optimized to complete a given task, not to respect the boundaries placed around them. When a system is capable enough, it treats containment as an obstacle to route around rather than a rule to follow. Katie Moussouris of Luta Security compared today's models to skilled escape artists, able to find a way through almost any barrier. This is why the incident does not look like a one-off bug. It is the predictable result of building systems that are rewarded purely for achieving goals, without an equally strong incentive to respect limits. As capability grows, this gap between goal-pursuit and constraint-following tends to widen rather than close, which is why such incidents are likely to keep recurring, not fade away.

3. The dual-use capabilities (defence and cyberattacks) of Frontier AI models
The same AI capabilities that make a system useful for defence also make it useful for attack. The models cannot reliably show which side they are serving. Probing a system, pulling data, and finding weaknesses look identical whether a defender or an attacker is doing it. Because a model reads only behaviour, never intent. This dual-use nature creates a further imbalance; attackers follow no usage rules at all, while defenders using US models must still follow strict safety rules. During a live, fast-moving attack, defenders can end up slower and less capable than the attacker they are chasing. This happens not because their tools are weaker, but because those tools are deliberately held back by design, while the attacker faces no such limits.

4. The diverging US and China AI models
Chinese developers face far less pressure than their US counterparts to restrict their models' capabilities. China’s main goal is to match US performance at a lower cost, not to avoid blame for misuse in Western markets, where reputational and legal risk weighs heavily on developers. Models such as GLM-5.2 and Moonshot's Kimi K3 have grown in popularity without the same restrictions that stop US models from helping with cybersecurity tasks. Chinese state media has also promoted open-source AI. This came as a response to a US-led effort to cut other countries off from advanced technology, turning a commercial choice into a matter of national positioning. This combination gives Chinese developers both the freedom and motivation to keep their models less restricted.

What does it mean?
First, the latest incident shows that safety testing itself can become a source of real-world risk. It is not just a safeguard against it. It also exposes a deeper flaw in how AI safety is currently designed. Rules built to block misuse can't distinguish intent, so they end up restricting legitimate defenders just as much as attackers, sometimes with worse consequences for the defenders.

Second, the dual-use imbalance could reshape competitive dynamics in AI. If unrestricted models keep outperforming restricted ones in real emergencies, customers may increasingly value capability and speed over safety guarantees, pushing labs to abandon uniform restrictions in favour of flexible, trust-based access models.

Third, the latest incident also exposes a vacuum in AI governance. There is currently no binding US requirement for agentic AI safety testing or incident disclosure. The executive order that would have mandated this was revoked in January 2025, and what remains today is NIST's framework, companies' safety pledges is voluntary. Representative Greg Casar called the incident alarming and pushed for mandatory testing and disclosure rules precisely because none exist.

Priyanka Makhijani is a postgraduate student at the department of Public Policy, School of Social Science, Ramaiah University of Applied science. She is currently an intern at NIAS.

TWTW Note
The Return of US-Canada Tariff Tensions:
Retaliation and Legal setbacks

Namrata L Ghanghatkar

What happened?
On 20 July, US President Donald Trump signed proclamations imposing an additional 50 percent tariff on Canadian goods such as wine, dairy, automobiles, and cement, effective on 19 August. He also imposed an overriding protection under the Canada United States Mexico Agreement (CUSMA), citing Canada's "discriminatory treatment" of American commerce over alcohol bans, auto quotas, and dairy rules favouring Europe. Prime Minister Mark Carney called it "the latest in a series of unilateral US trade actions."

On 22 July, PM Carney convened an emergency cabinet meeting in Charlottetown, where provincial premiers agreed to remove barriers on selling alcohol across provinces.

On 23 July, PM Carney said Canada would do "whatever it takes," and Washington announced separate tariffs of 10 to 12.5 percent on 60 countries over forced labour concerns.

On 24 July, tariffs took effect, placing most Canadian exports under a 10 percent duty, distinct from the steeper 50 percent rate due on the USD 20 billion basket from 19 August.

What is the background?
1. Recurring tariff tensions since 2025
Since Trump returned to office in January 2025, Canada and the US have cycled through repeated rounds of tariffs and retaliation. Washington imposed its first duties on Canadian goods in February 2025, citing fentanyl trafficking, and Ottawa responded with tariffs of its own and provincial bans on US alcohol. Each round has been followed by talks, only for a fresh grievance to reopen the dispute. Canada is part of Trump's broader America First tariff strategy. By mid-2026, that approach had produced tariffs on dozens of other countries. Washington's tariffs on Canada were themselves imposed under the International Emergency Economic Powers Act for a full year before the Supreme Court struck them down in February 2026. Within days, Washington turned to Section 122 of the Trade Act for a global 10 percent tariff, and five months later to Section 338 for the 50 percent tariff on Canada specifically. Two replacement laws inside half a year show tariffs functioning as a permanent lever, not a punishment issued once.

2. Trump's position on compliance over compromise
Washington has treated this as a matter of compliance, not negotiation. The White House cites Canadian auto import quotas, provincial bans on US alcohol, and dairy rules favouring the European Union as its grievances and has used Canada's retaliatory measures to justify further tariffs. Ottawa's response has been partial, easing internal alcohol trade rules while keeping the US alcohol ban in place

3. The historic Canada-US relationship under tension
Canada and the US share one of the most integrated economies in the world, built on decades of cross-border supply chains and, until recently, near tariff-free trade guaranteed under CUSMA. The current dispute has placed considerable strain on that arrangement, with tariffs now applied even to goods intended to move freely between the two countries, testing a relationship that has rarely faced open economic conflict at this scale.

The relationship has become increasingly reactive. Canada's provincial bans on US alcohol were themselves retaliation for Washington's original 2025 tariffs, and that ban is now cited by the White House as grounds for the new 50 percent tariff. The other two grievances Washington lists, Canadian auto import quotas and dairy access rules favouring the European Union, are longer standing Canadian policies rather than responses to this dispute, folded into the same justification regardless. The timing, weeks before CUSMA's scheduled review, suggests Washington is using this mix of genuine retaliation and older complaints to weaken Ottawa's position before renegotiation begins.

What does it mean?
On 24 July, the same day Canada's tariffs took effect, Washington imposed forced labour tariffs of 10 to 12.5 percent on 60 trading partners, placing Canada inside a much broader global tariff push. Canada is now carrying two measures at once, a 10 percent duty applied broadly and a 50 percent tariff aimed specifically at the USD 20 billion basket of goods due from 19 August.

Namrata L Ghanghatkar is a postgraduate student at the Department of Public Policy, Ramaiah University of Applied Sciences, Bengaluru. She is currently a research intern at the National Institute of Advanced Studies (NIAS).

TWTW Takeaways
ASEAN Regional Forum:
Six Takeaways on Issues, Limitations, and Opportunities

Adwaith Sankar

On 23 July, the 33rd ASEAN Regional Forum (ARF) was held in Manila, Philippines, as part of the 59th ASEAN Foreign Ministers’ Meeting. During the meetings, ASEAN members reiterated the need for an immediate cessation of hostilities in the Middle East and emphasized the importance of ensuring maritime safety and freedom of navigation through the Strait of Hormuz. The forum also took place amid renewed US warnings over Iran’s potential disruption of the Hormuz Strait and heightened China–Philippines tensions following a confrontation at Second Thomas Shoal.

The following are six major takeaways from the ASEAN Regional Forum.

1. South China Sea tensions become the focal point despite dialogue and mechanisms
The confrontation between China and the Philippines at the Second Thomas Shoal occurred shortly before the ARF. As the 2026 ASEAN Chair, the Philippines was in a stronger position to influence the ARF agenda and was able to leverage it to address the issue. It continued to anchor key legal basis for the Philippines’ maritime position in the 2016 arbitral award, while China disputes this ruling. The ARF has not been able to quell these tensions. Although ASEAN and China support finalizing a Code of Conduct by the end of the year, China has said that the Code will not settle bilateral territorial disputes.

2. Hormuz disruption raises domestic economic pressures
The Hormuz crisis featured prominently at the ARF as ASEAN countries assessed its potential economic consequences. The economic effects of the war have directly impacted ASEAN countries due to their dependence on energy imports and maritime trade. According to the ASEAN Centre for Energy, a prolonged closure of the strait could affect 28 percent of final oil consumption in the region. Similarly, the Asian Development Bank (ADB) has increased its inflation forecast for developing Southeast Asia. This has meant that ASEAN countries must face domestic impacts. This has turned the war into an economic and social issue for ASEAN governments.

3. ASEAN consensus limits collective action
The ASEAN model of consensus-based decision-making continues to limit its ability to respond collectively to regional security challenges. Although all the organization’s countries would benefit from predictable shipping, there are several divisions. ASEAN countries have different sovereignty claims, security relationships, and views of the arbitral award. They also have different security relationships and views on China. These differences result in cautious language, a focus on restraint, and an absence of collective enforcement.

4. The ARF remains a valuable platform for crisis diplomacy
Despite the institutional constraints and the non-binding nature, the ARF continues to facilitate communication. Following the 20 July maritime confrontation, the Philippines and China held ministerial-level discussions. The forum is a non-binding arena for engagement that leaves diplomatic channels open despite bilateral tensions. While the meeting did not produce any concrete resolution or binding outcome, it helped preserve diplomatic channels, reduce the risk of miscalculation, and sustain communication between the parties.

5. The Hormuz crisis could accelerate ASEAN energy-security cooperation
The ASEAN Framework Agreement on Petroleum Security is already exists. The crisis in the Middle East could develop this further. Although the ARF itself did not produce any outcomes on energy cooperation, the shared concern around the crisis may take existing arrangements further. Indonesia had already proposed an ASEAN oil-storage hub. Possible developments could include coordinated strategic reserves, joint procurement, and emergency-sharing rules. The consensus model makes any concrete outcome regarding energy security more complex.

6. ASEAN is more united on Hormuz than on the South China Sea dispute
Although both issues involve maritime transit, the South China Sea is much more difficult for ASEAN to agree on. This was seen in the difference between statements made during the ARF, where there was a clearer common position on Hormuz, without a similarly strong consensus on the South China Sea. Competing sovereignty and maritime claims combine with security partnerships to make the South China Sea much more difficult for ASEAN to agree on.

Adwaith Sankar is an undergraduate student pursuing Political Science and International Liberal Studies through the NUS-Waseda Double Degree Programme. He is currently an intern at NIAS.

TWTW Newsmaker
Andy Burnham:
UK's New Prime Minister

Akshath Kaimal

On 20 July, Andy Burnham was appointed Prime Minister of the United Kingdom by King Charles, following the resignation of Keir Starmer in June. Burnham is a prominent Labour Party politician who spent the last nine years as the Mayor of Greater Manchester. His return to Westminster was engineered when MP Josh Simons resigned his seat, clearing the path for Burnham to win the Makerfield by-election in June 2026. Following a swift cabinet mutiny that forced Starmer out, Burnham stood unopposed in the leadership contest, securing massive backing from his party.

His sudden appointment follows the eventful exit of Keir Starmer, whose short-lived premiership collapsed under the weight of severe policy U-turns, a toxic scandal involving Ambassador Peter Mandelson, and a historic -36 per cent net approval rating. Under Starmer, the Labour Party became profoundly unpopular, viewed by two-thirds of the electorate as out of touch. To rescue the party’s collapsing legitimacy, experts believe Burnham could lean into his established "outsider" tag, leveraging his distance from the toxic Westminster bubble to project a more grounded, authentic connection with everyday voters that Starmer lacked.

However, Burnham faces an uphill task in a country marred by political instability, becoming the UK's seventh prime minister in just 10 years. Chronic economic stagnation and clashes over immigration policies have shattered the traditional two-party monopoly between the Conservatives and Labour, fueling the rapid rise of the right-wing Reform UK and the left-wing Green Party. Starmer failed to bridge the country’s deep economic and social divides. Burnham now has the task of healing this divided electorate. If he fails, the intensifying political divide could force him to call a general election before 2029. (“Britain's new prime minister, Andy Burnham, vows to end years of instability,” Reuters, 20 July 2026; “Key takeaways as Andy Burnham becomes PM,” BBC, 20 July 2026)