US Orders Anthropic to Block Foreign Access to Advanced AI Models
The Trump administration ordered Anthropic to restrict non-US citizens from accessing its Fable 5 and Mythos 5 AI models, citing national security concerns over potential jailbreak vulnerabilities. Anthropic complied by disabling both models for all customers, though it disputes the government’s rationale, calling the risk minor and common. The move marks an aggressive export control action, sparking European and UK concerns about technological sovereignty and raising legal questions under the Export Controls Reform Act.
Cross-source coverage
Wire timeline
Darktrace Supports Anthropic's Temporary AI Model Restrictions on Security Grounds
Darktrace's SVP of AI and security strategy, Nicole Carignan, endorsed Anthropic's brief restriction of its Mythos and Fable AI models in the UK, arguing it was a prudent security measure rather than a geopolitical power play. The restrictions, imposed after US officials raised national security concerns about the models' potential for weaponization, were lifted after Anthropic agreed to additional safeguards. Carignan advised organizations to diversify AI suppliers and consider sovereign AI approaches to build resilience. The incident has heightened UK fears of over-reliance on US AI systems, with legal experts warning that export controls now extend to AI models themselves, marking a material shift in policy.
City AMTrump Administration Lifts Ban on Anthropic's AI Models After Security Fix
The Trump administration lifted its mid-June ban on non-Americans using Anthropic's latest Fable and Mythos AI models. The ban was initially imposed due to a potential breach of security protocols in the Fable model, prompting Anthropic to shut down all access to both models, including internally. The AI company subsequently implemented new safeguards and collaborated with the U.S. Commerce Department to resolve the security concerns. The incident underscores the ease with which the American government can restrict access to frontier AI technologies. Anthropic stated the episode highlights the industry's urgent need for a consistent framework to assess and mitigate potential evasions of AI safety measures.
Latest UpdatesTrump administration's embargo on Anthropic models undercuts American AI export strategy
The Trump administration's directive barring foreign nationals from accessing Anthropic's Mythos 5 and Fable 5 AI models has contradicted its own AI Action Plan aimed at promoting American AI exports worldwide. The article, written by former acting U.S. Commerce Secretary John Hannigan, argues that the cutoff intensifies global distrust of U.S. technology and accelerates 'AI sovereignty' initiatives across 46 countries investing in 135 sovereign AI projects. Drawing parallels to the Snowden/PRISM scandal, Hannigan warns that such actions brand allies as adversaries, damaging long-term commercial and diplomatic relationships. He recalls how the Obama administration's response to Snowden eventually led to transparency reforms and new data transfer frameworks. The piece calls on the Trump administration to adopt a more cooperative approach to manage global AI aspirations, rather than fracturing alliances through restrictive embargoes.
BrookingsUS lifts export controls on Anthropic’s Fable, Mythos AI models
The United States has lifted export controls on Anthropic's Fable and Mythos AI models, allowing the company to release its Claude Mythos 5 artificial intelligence model to select trusted US organizations. This decision comes as Washington increases oversight of new AI model releases to identify potential threats posed by advanced AI systems. The move signals a shift in US policy towards balancing national security concerns with promoting domestic AI innovation and competitiveness.
The Business TimesUS lifts export controls on Anthropic’s Fable, Mythos AI models
The United States government has lifted export controls on Anthropic's Fable and Mythos AI models, allowing the company to release its advanced Claude Mythos 5 artificial intelligence model to some US organizations. The decision, announced by Anthropic on June 26, 2026, comes as Washington has stepped up oversight of new model releases to identify potential threats posed by advanced AI systems. The move signals a shift in US regulatory approach, balancing national security concerns with the promotion of domestic AI innovation. The article, published by The Business Times Singapore on July 1, 2026, highlights the ongoing tension between AI development and export control policies.
The Business TimesAnthropic's Export Control Crackdown Leaves South Korea Caught in Washington's AI Crossfire
The U.S. Trump administration ordered Anthropic to suspend access to its advanced AI models (Fable 5 and Mythos 5) worldwide on June 12, 2025, following alleged security incidents involving SK Telecom and a reported jailbreak. This came days after SK Telecom and Nvidia announced a major AI cloud partnership in South Korea. The article reveals how Anthropic, which had spent nearly $5 million lobbying for stricter AI chip export controls, now faces the same logic applied against its own software. Critics like Chatham House describe U.S. policy as inconsistent and mercurial. For South Korea, the move threatens the viability of its AI infrastructure bet on U.S. technology, as Chinese markets are increasingly closed to Korean firms. The episode highlights how corporate rivalries, not grand strategy, shape U.S. export decisions.
The DiplomatAnthropic Export Controls Strain US-South Korea AI Relations
The Trump administration's June 12 export control directive forced Anthropic to suspend access to its advanced AI models, Fable 5 and Mythos 5, globally after alleged security incidents involving South Korea's SK Telecom. The move came days after SK Telecom and Nvidia announced a major AI cloud partnership, exposing a clash between Silicon Valley lobbying and allied economic interests. Anthropic, which had previously advocated for stricter U.S. AI chip export controls, faced backlash as its own software was restricted. The incident highlights the inconsistency and unpredictability of U.S. AI export policy, which critics say harms allied nations like South Korea. SK Telecom, with minimal Chinese exposure, was excluded from access to Mythos over unverified allegations, jeopardizing its multi-billion-dollar AI infrastructure bet on the U.S. ecosystem. Meanwhile, China's $295 billion plan to build a domestic AI supply chain further pressures Korean firms reliant on U.S. technology.
The DiplomatAnthropic asked for regulation. Washington went much further.
Anthropic, the AI company valued at nearly $1 trillion, has long advocated for government regulation of artificial intelligence. However, the Trump administration issued an export control directive ordering Anthropic to suspend access to its latest Claude models, Fable 5 and Mythos 5, by any foreign national, including its own foreign employees. The administration cited national security concerns but did not specify details. Anthropic CEO Dario Amodei had recently published an essay calling for binding AI regulation, including the ability to block unsafe models. The company disagreed with the administration's finding, calling it a misunderstanding and arguing the action lacked transparency and fairness. Senior Anthropic employees traveled to Washington D.C. to meet with administration officials. This marks the second time in 2026 that Anthropic has faced government action, following a previous Pentagon-related ban.
US Top News and AnalysisAnthropic Shutdown: In the Beer Garden with Superintelligence
The article reports on AI company Anthropic's decision to take its newest Claude models, Mythos 5 and Fable 5, offline on June 12, 2026, following a US government order to block non-American access. Anthropic warned about the dangers of its own products, citing security concerns related to the models' ability to identify software vulnerabilities, which could be exploited by malicious actors. The column contrasts Anthropic's cautious, ethics-focused approach with rival OpenAI's more permissive stance, and explores the tension between altruistic motives and profit-driven business goals, as both companies plan IPOs. The author reflects on conversations with tech-savvy friends about AI's transformative potential and risks, questioning who ultimately controls and decides the technology's deployment.
taz.de - taz.deWhite House Blocks Anthropic's New AI Models Over National Security Concerns
The Trump administration has effectively blocked public access to Anthropic's latest AI models, Claude Mythos 5 and Claude Fable 5, citing national security concerns. The White House issued a directive under export control rules, barring use by foreign nationals in the U.S. or anyone abroad. Anthropic temporarily suspended access, stating the government cited a potential jailbreak vulnerability, which the company describes as minor and not unique to its models. The move follows months of tension after Anthropic refused to allow its technology to be used for mass surveillance or autonomous weapons, leading to a 'supply chain risk' designation and a lawsuit. A federal judge temporarily blocked that designation, calling it 'corporate murder.' Anthropic argues the action lacks transparency and fairness, and if applied industry-wide, would halt all new model deployments.
Reason.comBlocking of Anthropic's AI Models: A Well-Timed Warning for Europe
The US government has ordered California-based AI company Anthropic to block its powerful AI models 'Fable 5' and 'Mythos 5' for foreign users and employees, citing national security concerns. Anthropic argued it could not distinguish between domestic and foreign users and subsequently blocked access for everyone. While the immediate impact on Europe is minimal due to limited availability, the commentary warns that this decision should serve as a wake-up call for European companies and authorities relying on US tech products like Microsoft, Google, Amazon, or Palantir. The author argues that the timing is optimal to push for European self-reliance in AI, as the uncertainty of service availability undermines trust in US AI models. The piece emphasizes the need for Europe to prepare for potential future blocks and reduce dependency on US technology.
taz.de - taz.deLegal Considerations Related to the Anthropic Export Controls Directive
On June 12, Anthropic disabled access to its advanced AI models Claude Fable 5 and Mythos 5 after receiving an undisclosed order from the U.S. government, citing national security authorities and export controls. The order, reportedly issued by Commerce Secretary Howard Lutnick, bars any non-U.S. national from accessing these models, even within the United States. Legal experts argue the directive is unprecedented in breadth, contrasting sharply with the Trump Administration's prior deregulatory AI stance. The legal basis is likely the Export Controls Reform Act of 2018, though the government has not publicly justified the order. Speculation grows that Anthropic is being singled out for scrutiny, potentially linked to its designation as a supply chain risk by the Department of War. The article calls for public transparency from the Commerce Department regarding the nature of the threat and the likelihood of similar restrictions on other AI models.
Just SecurityLegal Considerations Related to the Anthropic Export Controls Directive
On June 12, Anthropic disabled access to its Claude Fable 5 and Mythos 5 AI models after receiving an unprecedented U.S. government export controls directive. The order, reportedly from Commerce Secretary Howard Lutnick, cites national security concerns about jailbreaking vulnerabilities, though Anthropic describes these as minor and common. The legal authority is likely the Export Controls Reform Act of 2018, used previously for semiconductor exports but never applied so broadly to AI models. The directive restricts access by any non-U.S. national worldwide, marking a sharp reversal from the Trump Administration's deregulatory AI stance. Commerce has not publicly disclosed the order or its rationale, leading to speculation about targeted scrutiny of Anthropic following its designation as a supply chain risk by the Department of War. The article calls for public guidance on the threat and broader AI safety regulations.
Just SecurityAnthropic Suspends Top AI Models After US Export-Control Order
The Trump administration ordered Anthropic to restrict foreign-national access to its two most advanced AI models, Fable 5 and Mythos 5, citing national security concerns over a potential jailbreak. Anthropic responded by disabling both models for all customers while it works to comply, though it disputes the government's rationale, arguing the alleged vulnerability is narrow and already available from other models like OpenAI's GPT-5.5. The order marks an aggressive step in controlling frontier AI with dual-use cybersecurity capabilities. The Department of Defense's CIO publicly supported the move, emphasizing national security over commercial interests. The shutdown complicates deployment plans for federal agencies and critical infrastructure partners, and raises unresolved questions about balancing trusted access with preventing misuse by adversaries.
Defense One - All ContentTrump Ban on AI Access to Foreign Users Forces Anthropic to Suspend Models
President Trump has ordered Anthropic, the maker of the Claude AI model, to ban non-American nationals from accessing its newest AI models, Fable 5 and Mythos 5, citing national security concerns over potential jailbreak vulnerabilities. Anthropic complied with the legal order but disagreed with the decision, warning that consistent application of such standards could lead to more bans. The move has sparked concern among UK politicians and analysts about Britain's ability to maintain technological sovereignty and keep pace with AI advancements. UK AI Minister Kanishka Narayan defended the government's investment in sovereign AI capabilities, while former security minister Tom Tugendhat and others criticized the UK's response as insufficient. The ban follows previous tensions between the Trump administration and Anthropic over the company's ethical restrictions on military use of its AI.
City AMAnthropic Blocks All Access to Cyber AI Models After US Order to Exclude Non-Citizens
Anthropic announced it has suspended all access to its advanced cyber-capable AI models, Mythos 5 and Fable 5, following a US government order to exclude non-US citizens from using them. The company stated it disagrees with the decision and is working to restore access. The US government acted after being alerted to potential security bypass vulnerabilities in Fable 5. The move has reignited European calls for sovereign AI cyber capabilities, as the EU's cybersecurity agency ENISA was granted access in June after negotiations. Anthropic believes the government's concerns are unwarranted, arguing the model's capabilities are in line with others on the market. The incident highlights growing tensions between national security, AI governance, and global access to cutting-edge technology.
EuractivAnthropic Disables Fable 5 and Mythos 5 AI Models to Comply with US Government Export Control Directive
Anthropic announced it disabled access to its Fable 5 and Mythos 5 AI models to comply with a US government export control directive citing national security authorities. The company received an order on Friday at 5:21 PM ET instructing it to suspend all access to the models by any foreign national, including foreign national Anthropic employees, both inside and outside the United States. To ensure compliance, Anthropic abruptly disabled the models for all customers, though other models remain unaffected. The move comes days after Anthropic released Fable 5 and Mythos 5, touting them as state-of-the-art across industry benchmarks. Fable 5 marked the first public release of such an advanced offering with new safeguards. The government did not provide specific details about its national security concern. Anthropic criticized the action as lacking transparency and fairness, noting it believes the government should have the ability to block unsafe deployments through a statutory process grounded in technical facts. This incident follows a previous clash with the Department of Defense, which declared Anthropic a supply chain risk, leading to ongoing litigation.
US Top News and Analysis