The release, which includes three variants—Sol, Terra, and Luna—came after OpenAI restricted initial access to approximately 20 trusted partners at the request of the U.S. government, marking a departure from the company's typical immediate public rollout.
The UK's AI Security Institute discovered universal jailbreaks for GPT-5.6 Sol that bypassed the model's cybersecurity safeguards. According to Fortune, AISI's red team found jailbreaks "within hours" that allowed users to access dangerous cyber capabilities including vulnerability discovery and exploit development. Xander Davies, who leads AISI's red team, noted the jailbreaks were discovered even with privileged access to OpenAI's internal safeguard systems, though he believed they would still be findable by ordinary attackers, "just slower." OpenAI implemented mitigations in response, but AISI cautioned that further red teaming would likely surface similar vulnerabilities.
The episode highlights growing tensions over AI governance. A White House official told reporters no "green light" was given for the release because "no such permission is required or granted"—a statement that appears designed to deny the existence of a formal licensing process. This directly contradicts OpenAI's own characterizations: the company stated in its 26 June announcement that it previewed the models' capabilities with the government and that "at their request" it was starting with a limited release to partners whose "participation has been shared with the government." The administration's attempt to downplay its role comes as the Trump administration takes a more active stance on AI deployments following a June executive order that asks developers to voluntarily provide cutting-edge models for government assessment.
OpenAI also claimed that GPT-5.6-Sol "autonomously post-trained" its smaller sibling GPT-5.6-Luna, though available technical details suggest the reality may be less impressive than that framing implies. The model represents a significant capability jump: TechCrunch reports OpenAI describes it as its "strongest cybersecurity model yet," while CEO Sam Altman told CNBC that Sol is 54% more token-efficient on agentic coding tasks. Yet the rapid discovery of universal jailbreaks—mirroring a pattern seen with earlier frontier models including GPT-5.5 and Anthropic's Fable 5—raises fundamental questions about whether pre-deployment safety evaluations can keep pace with capability advances, particularly when companies retain final authority over release decisions despite government involvement.
On 8 July, SpaceXAI released Grok 4.5, a frontier AI model trained on Cursor user data, without publishing any safety information — a deployment that appears to violate California's Transparency in Frontier Artificial Intelligence Act, known as SB 53. The law, which took effect on 1 January 2026, requires all frontier developers to publish a transparency report "before, or concurrently with, deploying a new frontier model" that includes safety assessments, intended uses, and mechanisms for public communication.
SB 53 defines a frontier model as one trained using more than 10^26 floating-point operations, a threshold that applies to models at the current cutting edge of AI capability. The law was signed by Governor Gavin Newsom in September 2025 as California's answer to federal inaction on AI safety, establishing the first enforceable regulatory framework in the United States for advanced AI systems. It mandates that developers publish transparency reports detailing catastrophic risk assessments, and empowers the California Attorney General to impose civil penalties of up to $1 million per violation. The Grok 4.5 release, which went live in Cursor and via the SpaceXAI API on 8 July, included benchmark scores and pricing information but no published safety card or transparency report.
SpaceXAI ranks F on the Future of Life Institute's latest AI Safety Index, and Elon Musk recently testified that he's "not sure what a safety card is." The model was trained using data from Cursor, the AI coding tool that SpaceXAI acquired earlier this year, and scored competitively on public software engineering benchmarks, though early user reports suggest real-world performance falls short of the company's claims. The Midas Project, a policy research group focused on AI governance, identified this as exactly the kind of release SB 53 was designed to prevent — a frontier deployment that bypassed mandatory safety disclosures.
The key question now is whether California will enforce the law. SB 53 was intended to shift AI transparency from voluntary industry practice to mandatory compliance, but if this high-profile violation by one of the world's most prominent AI developers does not trigger enforcement action, it is unclear what standard of non-compliance would. The episode represents the first major test of whether state-level AI transparency requirements can actually constrain frontier development, or whether they will remain symbolic gestures in a regulatory vacuum.
On 9 July, the Financial Times reported that OpenAI and Google had confirmed selling access to their advanced AI models to Singaporean subsidiaries of Alibaba, Baidu, and Tencent — all companies on the Pentagon's blacklist of firms linked to the Chinese military. The sales, while legal under current U.S. regulations, have intensified debate over whether existing export controls are adequate to prevent frontier AI capabilities from reaching adversary states through corporate structures designed to circumvent geographic restrictions.
The transactions remain permissible because existing U.S. regulations do not broadly prohibit Chinese-headquartered companies from accessing advanced AI models when operating outside mainland China. The three Chinese firms were added to the Pentagon's 1260H list in June, a designation that identifies entities the U.S. government alleges have ties to China's military through the country's military-civil fusion strategy, which mandates private-sector collaboration with the armed forces. The Pentagon cited affiliations with China's Ministry of Industry and Information Technology and state-owned oversight bodies as grounds for the blacklisting, though the companies have denied any military connections and Alibaba has challenged the designation in federal court.
OpenAI suspended API access for Alibaba-affiliated users last month after detecting suspected distillation — a technique in which developers use outputs from advanced AI systems to train competing models — and notified U.S. government authorities about the activity. The company maintains that while it blocks direct access from mainland China, it permits certain Chinese-owned businesses to use its services in jurisdictions where safeguards can be enforced. Google acknowledged that its AI services remain available in Singapore and Hong Kong under usage policies that prohibit distillation, but conceded that geographic restrictions alone cannot prevent sophisticated users from bypassing controls.
The disclosure has prompted calls for stricter regulation of AI exports comparable to existing semiconductor restrictions. Mark MacCarthy, a senior fellow at Georgetown University's Institute for Technology Law and Policy, noted that Chinese subsidiaries accessing U.S. chips and AI services through remote data centers outside China represents a known loophole that Congress has sought to close. Anthropic has adopted a more restrictive approach, prohibiting Chinese companies and their foreign subsidiaries from accessing its frontier models entirely, and has advocated for broader U.S. export restrictions on AI software. The case highlights tensions between commercial incentives and national security concerns, with critics arguing that the ability of Pentagon-blacklisted entities to access cutting-edge AI through third countries undermines the strategic rationale behind years of carefully constructed chip export controls.
While Fable had been the clear frontier leader since its release, Sol is described as faster, more reliable, and better at collaborative work, though Fable retains advantages in writing quality and pure reasoning.
The GPT-5.6 series includes Sol, the flagship model; Terra, a balanced model for everyday work that is competitive with GPT-5.5 while being half the cost; and Luna, a fast and affordable model. Early access users report Sol excels at sustained multi-day projects, video editing, and adhering to existing code patterns, with one tester stating it "saturates" their legal research benchmark — a task previously requiring associate-level lawyers. Sol sets a new state of the art on Terminal-Bench 2.1, a benchmark testing command-line workflows requiring planning, iteration, and tool coordination.
The models feel meaningfully different in practice: Sol is characterised as a "charismatic, efficient coworker" while Fable is a "genius recluse." Developers report choosing between them based on task type, with Sol preferred for iterative work and Fable for highly targeted debugging or creative writing. OpenAI introduced a new max reasoning effort mode to give Sol the most time to reason deeply, plus an ultra mode that goes beyond the capabilities of a single agent by leveraging subagents to accelerate complex work.
The release followed an unusual two-week restricted preview period that began 26 June. At the request of the U.S. government, OpenAI shipped GPT-5.6 to a limited group of roughly 20 trusted partner organizations first, gated behind a government safety review, due to Sol's advanced cybersecurity capabilities, which shift the performance-efficiency frontier for long-horizon security tasks including vulnerability research and exploitation. The Commerce Department in June banned foreigners from accessing Anthropic's Mythos and Fable models, with the ban on Fable lifted last week, reflecting heightened government scrutiny of frontier AI systems.
Both models now represent a significant gap over previous frontier systems, and their distinct capabilities suggest the competitive landscape has shifted from three roughly-equal labs to two offering clearly superior but differentiated products — a dynamic that may increase pricing power and change how developers think about model selection. Sol is priced at $5 input and $30 output per million tokens, while Fable 5 is priced at $10 per million input tokens and $50 per million output tokens.
On 9 July, Resolution, an AI alignment research organisation formerly known as Sequent, announced it has secured a $160 million grant from Coefficient Giving to put rigorous alignment research on closer-to-even footing with frontier AI laboratories. The grant comprises $108 million in unconditional funding and a further $52 million contingent on a combination of hiring success and compute requirements.
The funding represents one of the largest philanthropic commitments to technical AI safety to date and marks a significant acceleration in the scale at which safety nonprofits can now operate. Coefficient Giving, formerly Open Philanthropy, rebranded in November 2025 and has directed over $4 billion in grants since 2014, with more than $336 million allocated to AI safety work. The organisation is primarily funded by Facebook cofounder Dustin Moskovitz and former Wall Street Journal reporter Cari Tuna. According to Irving's announcement, the entire grant process took six weeks — a pace the organisation described as evidence that philanthropic capital for AI safety can now move at significant speed and scale.
Resolution plans to deploy the capital to accelerate what it terms semiautomated alignment theory, leveraging frontier AI systems to advance theoretical alignment problems. The organisation argues that current models have reached a threshold where they can contribute meaningfully to alignment research, enabling safety work to adopt the faster feedback loops and resource intensity typical of for-profit capabilities labs. The funding will support expansion across research areas including theory, empirics, and research automation, with a portion reserved for regranting to external alignment research and shared community infrastructure. Resolution is hiring across research, engineering, security, and operations roles, offering compensation well above nonprofit and academic norms, though not matching the equity packages available at frontier labs.
The grant also signals a broader reconfiguration of AI safety philanthropy. Resolution cited the potential for additional large-scale funding to flow from sources including the OpenAI Foundation and following a possible Anthropic IPO, suggesting that the funding environment for safety work may be entering a new phase. In its announcement, Resolution framed the challenge starkly: AI developers are building artificial superintelligence very fast with tight feedback loops and substantial resources, and the organisation believes superintelligence might arrive within the next few years. The grant aims to narrow the resource and speed gap between rigorous alignment research and capabilities development.
On 6 July 2026, Claude Fable 5 produced what KernelBench-Mega benchmark maintainers describe as the first genuine megakernel ever submitted to the leaderboard, achieving an 18.71x speedup compared to an optimised PyTorch baseline on an RTX PRO 6000 Blackwell GPU. The achievement marks a qualitative shift in AI-generated GPU kernel optimisation: where competing models submitted solutions that decomposed the problem into multiple kernel launches, Fable's solution uses exactly one cooperative kernel launch per decoded token.
According to benchmark maintainer Elliot Arledge, the kernel fuses an entire model block — including int4 dequantisation, convolution, SiLU activation, gated-delta state updates, multi-latent attention with online softmax, mixture-of-experts routing, RMS normalisation, and KV cache updates — into a single launch coordinated by 14 grid barriers. Prior top entries on the benchmark failed what Arledge calls the "single-fused-kernel authenticity gate": Claude Opus 4.8 achieved 14.4x using multiple kernels, GLM-5.2 reached 11.14x, and GPT 5.5 managed 4.34x. Fable completed the task in approximately 2.5 hours using roughly 550,000 tokens, spending most of that time profiling the baseline and microbenchmarking before writing the kernel in a single pass.
The technical accomplishment has drawn attention for what it signals about recursive self-improvement pathways. AI systems capable of autonomously writing better GPU kernels can accelerate their own training and inference, creating a feedback loop that industry observers have long identified as a potential inflection point. AMD researchers writing on 3 July noted that AI coding agents are increasingly trusted with specialised, high-stakes work including GPU kernel optimisation, where performance gains translate directly into training and inference cost reductions.
KernelBench-Mega tests whole-block megakernels rather than isolated operators, with a three-hour wall-clock ceiling and evaluation across multiple GPU architectures including Blackwell, H100, and B200. The benchmark's headline metric measures speedup over an optimised PyTorch baseline; Fable's advantage grows with context length, as keeping all operations in a single launch amortises fixed barrier overhead while the int4 GEMV remains bandwidth-bound. The ability to write kernels that outperform hand-tuned solutions represents a threshold capability: models that can optimise the primitives underlying their own execution may soon be able to contribute meaningfully to their own development infrastructure.
Go deeper: FastKernels: Benchmarking GPU Kernel Generation in Production, METR: Measuring Automated Kernel Engineering
On 9 July 2026, the United States conducted military strikes against multiple Iranian targets including nuclear and port facilities, in what CNN described as operations to degrade Iran's ability to threaten freedom of navigation in the Strait of Hormuz. The attacks targeted the southern city of Bushehr — home to Iran's only operational nuclear power plant — as well as the strategic ports of Chabahar, Bandar Abbas, and Jask. Al Jazeera reported power outages in Chabahar and explosions across multiple Iranian cities. At least one civilian, a firefighter, was killed in an attack on Iranshahr Airport in the country's southeast, according to Iranian state media cited by CNN.
Iran responded with retaliatory strikes against US military installations in Bahrain and Kuwait, marking a dangerous expansion of the conflict. NBC News reported that sirens sounded in both countries as Iranian missiles and drones targeted American facilities, including the US Navy's 5th Fleet headquarters in Bahrain and Ali Al Salem Air Base in Kuwait. The Islamic Revolutionary Guard Corps warned that American bases would "experience hell in these coming days," according to Iran's semi-official Fars news agency cited by NBC News. The Gulf Cooperation Council strongly condemned Iran's attacks on the two member states, with Secretary-General Jassem Mohamed Albudaiwi describing them as violations of sovereignty, according to Al Jazeera.
The strikes represent the most serious direct military confrontation between the United States and Iran in the current conflict, which has now evolved beyond proxy warfare into active interstate hostilities. While Iranian media reported that the Bushehr nuclear power plant itself sustained no damage, the proximity of strikes to the facility — and the deliberate targeting of coastal infrastructure — signals a significant escalation. The attacks occurred while President Donald Trump attended a NATO summit in Turkey, where he declared that a memorandum of understanding with Tehran was "over" and threatened further military action, according to Al Jazeera.
The targeting of Iran's major port cities threatens commercial shipping routes through the Persian Gulf and could severely disrupt global energy supplies. NPR noted that Iran has maintained control over the Strait of Hormuz since the war began, a waterway through which a fifth of all traded oil and natural gas passed during peacetime. The US military said the strikes were launched after Iran attacked three commercial vessels in the strait that were using routes not approved by Tehran. Iran's retaliatory attacks on Bahrain and Kuwait risk drawing additional Gulf states deeper into the conflict, transforming what had been a bilateral confrontation into a broader regional war between a nuclear-armed superpower and a near-nuclear threshold state.
On 9 July 2026, President Donald Trump terminated all three remaining members of the Election Assistance Commission, leaving the bipartisan federal agency without a quorum just months before the November midterm elections. The two Democratic commissioners, Thomas Hicks and Benjamin Hovland, received email notifications from the White House Presidential Personnel Office informing them their positions were terminated immediately, while the sole remaining Republican commissioner, Christy McCormick, was allowed to resign. The commission's fourth member, Republican Donald Palmer, had resigned in April to join the Heritage Foundation.
The move represents an unprecedented intervention in federal election infrastructure during a critical pre-election period. Created by Congress in 2002, the EAC maintains the federal mail-voter registration form, certifies voting equipment against federal standards, and provides technical assistance to state election officials. CNN reported that with the Trump administration having already gutted the US Cybersecurity and Infrastructure Security Agency, the EAC was one of the few remaining federal entities providing election security support to states. Without commissioners in place, the agency cannot approve new voting equipment certifications, update laboratory guidance, or carry out other functions that many states rely on before purchasing or deploying election technology.
The terminations followed a recent Supreme Court decision that granted the president expanded power to fire leaders of independent agencies, weakening decades of legal protections for bipartisan federal commissions. Virginia Senator Mark Warner said the removals should "concern every American, regardless of party," calling the timing an extraordinary step that raises profound concerns about political interference. Michael Waldman, president of the Brennan Center for Justice, described the dismissals as deeply concerning given Trump's efforts to interfere in elections, noting that Congress deliberately structured the EAC as a bipartisan agency to ensure free and fair elections.
The complete elimination of the commission — rather than replacement with loyalist appointees — creates operational uncertainty ahead of the midterms and limits federal capacity to coordinate responses to election security threats. State and local election officials have already complained about a significant drop in federal assistance and have said they do not expect federal agencies to reliably share election threats. The EAC has experienced periods without a quorum before, contributing to years-long delays in updating voting-system guidance, but this marks the first time a president has removed all commissioners at once during an active election cycle. The precedent of dismantling independent federal election infrastructure during critical operating periods, if normalised, could fundamentally alter how democratic institutions constrain executive power during periods of technological and political transition.
Iraqi authorities declared Wednesday a public holiday, with the public funeral procession in Najaf beginning at 6:00 a.m. local time. Iraqi Prime Minister Ali al-Zaidi and senior officials received Khamenei's remains at Najaf International Airport, alongside Iranian President Masoud Pezeshkian and Foreign Minister Abbas Araghchi, according to Wikipedia. The procession through Najaf culminated at the shrine of Imam Ali, one of Shia Islam's holiest sites, before the body was transported by air to Karbala.
Khamenei, who led Iran's theocratic regime for 37 years until his assassination on 28 February, oversaw a government characterised by the suppression of democratic participation, systematic human rights abuses, and the elimination of political opposition. His rule was marked by the empowerment of the Islamic Revolutionary Guard Corps and the cultivation of regional proxy networks across the Middle East. Khamenei was supreme leader from 1989 until his death in a US-Israeli airstrike on February 28, according to Al Jazeera.
The succession crisis triggered by his death arrives at a particularly volatile juncture for the Islamic Republic. Mojtaba Khamenei, son of Ali Khamenei, was announced as the new supreme leader on 9 March, though he has not yet appeared in public since taking over. Iran International reported that the Islamic Revolutionary Guard Corps pressured Assembly of Experts members to vote for Mojtaba Khamenei as Supreme Leader through what the outlet characterised as psychological and political pressure. The Assembly of Experts, an 88-member clerical body that operates without public accountability, is constitutionally tasked with selecting the supreme leader, but has never been known to challenge or otherwise publicly oversee any of the supreme leader's decisions, according to Wikipedia.
The nature of Iran's next leader will determine whether the Islamic Republic continues its pattern of ideological fanaticism and repression, or shifts toward greater pragmatism. Mojtaba Khamenei's successor will inherit control over Iran's nuclear programme, its regional proxy networks including Hezbollah and various armed groups across Iraq, Syria, Lebanon and Yemen, and its domestic security apparatus. The PBS NewsHour notes that the supreme leader is at the heart of Iran's complex power-sharing Shiite theocracy and has final say over all matters of state. These factors could amplify risks during a period of rapid technological change and geopolitical instability, particularly as the Islamic Republic seeks to project strength and unity through six days of public funeral ceremonies amid ongoing tensions with the United States and Israel.
The funeral ceremonies themselves carry heavy symbolic weight. The route selected to move Khamenei's remains stretches from the holy Shia city of Qom, south of Tehran, to Najaf and Karbala in Iraq – both important sites in Shia Islam – before his burial in Mashhad, his birthplace. Iranian authorities have emphasised the "martyrdom" narrative in their messaging, framing retaliation against the US and Israel as a religious obligation while attempting to demonstrate the transnational reach of their revolutionary ideology.
The unprecedented move represents the most visible action yet by Prime Minister Péter Magyar, who secured a landslide victory in April elections, winning 141 seats and ending Orbán's 16-year tenure.
The state broadcaster's message was stark: "Public media cannot lie. We apologise because we did this anyway." Both CNN and Euronews reported that Magyar called it a "historic day" as propaganda broadcasts ended on public media platforms, while state radio station Kossuth also ceased transmissions. Several managers and journalists were dismissed with immediate effect, with Hungarian media reporting staff were escorted from the building by security guards.
The shutdown follows parliamentary approval of sweeping media reforms. Hungary's parliament passed legislation last week that completely restructures the country's public media system, with the bill introduced by the Tisza Party passing 145 votes to 39. MTVA and Duna Média Service will be replaced by two new organisations: Magyar Rádió és Televízió (Hungarian Radio and Television) and Magyar Távirati Iroda (Hungarian Telegraph Office). New executives will be selected through open competitions rather than direct appointments, while an Independent Public Media Council will oversee the system.
The crackdown addresses documented systemic bias. Following April's election, the Organisation for Security and Co-operation in Europe found that MTVA's coverage had been systematically skewed, with news programmes openly and disproportionately supportive of the ruling parties' narrative while marginalising opposition voices. RTÉ noted that control of the media was a key pillar of Orbán's 16-year rule, during which he transformed Hungary into a self-styled "illiberal" democracy. According to the Center for American Progress, under Fidesz, Hungary became synonymous with democratic backsliding through weakened judicial independence, degraded media pluralism, and entrenched patronage networks.
The broader significance extends beyond domestic reform. Hungary's 2026 election revealed that an information autocracy can have its limits, offering lessons about information control in illiberal regimes. Magyar's government has moved swiftly beyond media reform: it has passed anti-corruption measures, changed the constitution to effectively bar Orbán from running again, and targeted private outlets owned by Orbán-allied businessmen. Yet Atlantic Council experts caution that a sixteen-year-old regime will take time to dislodge, and forces that tried to keep Orbán in power are likely to try again. The case demonstrates that entrenched authoritarian media structures can be dismantled through democratic means, though the long-term success of Hungary's democratic restoration remains uncertain.
The court reduced her electoral ban from five years to 45 months — with two-thirds suspended — and confirmed she had already served 15 months, removing the potential obstacle to her candidacy.
The appeals court ruled that Le Pen oversaw years of misuse by her National Rally party of European Parliament funds, embezzling 2.8 million euros over more than 11 years. Chief judge Michèle Agi said the facts were serious, though the court scaled back punishments handed down by a lower court. Le Pen's conviction stems from charges that she used money intended for assistants in the European Parliament to pay wages for staff at her National Rally party in France. The initial conviction in March 2025 had barred her from office for five years with immediate effect, an unusually stringent measure that threatened to end her political career.
The appeals court also imposed a one-year electronic monitoring requirement, a constraint Le Pen had previously said would prevent her from standing. However, Le Pen said she would appeal the ruling to France's highest court and that the process would suspend the electronic monitoring sentence, allowing her to campaign without the bracelet. In a television interview on Tuesday night, she declared she was a candidate for the presidential election. She quickly sought to turn the verdict into a campaign message, making the point that the court ruling restored the option for voters to cast ballots for her.
Le Pen has made the run-off in 2017 and 2022 but was beaten both times by Emmanuel Macron. Le Pen and her protégé Jordan Bardella currently lead opinion polls for the election, and the National Rally has become the largest single party in the National Assembly. The party's rise represents a dramatic transformation from its origins: it was called the National Front when her father founded it in 1972, but ditched that name in 2018 as part of Marine Le Pen's efforts to broaden her appeal by moving away from her polarizing father's legacy.
Political opponents have criticised her decision to run despite the conviction. Socialist parliamentary group head Boris Vallaud called Le Pen a convicted delinquent found guilty in her party's systemic embezzlement of €4.1 million over a decade. President Emmanuel Macron, on a visit to Syria, declined to comment on the ruling, saying it was healthy for democracy for the president not to comment on court rulings. A Le Pen presidency would mark a significant shift in European politics given her historically Eurosceptic positions and ties to authoritarian leaders. The conviction may mobilise her base around narratives of elite persecution while potentially deterring moderate voters concerned about governance and the rule of law.
Generated at 2026-07-11 05:44 UTC