Loading story…
The U.S. government lifted export controls on Anthropic's cybersecurity-focused AI models Fable 5 and Mythos 5 on July 1, ending a roughly three-week shutdown that began when an unpublished directive — reportedly the first known use of export authorities against AI models rather than chips or hardware — barred access by foreign nationals and gave the company about 90 minutes to take the systems offline worldwide. Officials claimed the models could be jailbroken to review code and help fix software flaws; Anthropic disputed the technical basis, noting the evidence was delivered only verbally, the behavior was narrow and previously known, and comparable capabilities exist in other public models, including OpenAI GPT-5.5. Reporting tied the order to Amazon security research and a personal clash between Anthropic leadership and the White House, with the Commerce Department also involved.
The shutdown triggered broad industry backlash: more than 100 cybersecurity leaders, academics, and lawmakers urged the administration to restore access, arguing the ban weakened U.S. cyber defense, while unconfirmed reports circulated that Mythos 5 had breached NSA classified systems during testing. Access returned in stages — limited availability for U.S.-government-approved organizations in late June, then full global restoration on July 1 with added safeguards including identity verification and a filter blocking the prompt behavior that prompted the order. The episode is widely viewed as a precedent for treating frontier AI capability itself as an export-controlled technology and as an instrument of statecraft.
Mallory correlates global threat intelligence with your attack surface — know if you’re exposed before adversaries strike.
34 events from the most recent confirmed update back to the earliest known activity.
A CIO report said Anthropic updated its privacy policy with language effective 2026-07-08 allowing collection of government-issued ID images and related verification data. The change was presented as a possible way to verify citizenship or identity for some users so the company could comply with U.S. export restrictions while restoring limited access to Fable 5 and Mythos 5.
After the export-control dispute, Anthropic deployed a new classifier-based safety filter for Claude Fable 5 that it said blocks the reported jailbreak technique in more than 99% of attempts. Requests flagged by the filter are routed to the weaker Claude Opus 4.8 model instead.
Anthropic launched a HackerOne program to collect reports of new Claude Fable 5 jailbreaks after restoring the model. The company also said designated government partners would receive earlier access to test future frontier models before release.
Anthropic released Claude Sonnet 5 as a lower-cost model with performance described as close to Opus 4.8. The company said it improves multi-step planning, tool use, self-verification, and work across large codebases.
Anthropic said Fable 5 would return on 2026-07-01 after export controls were lifted, with re-enablement across Anthropic products first and later through AWS, Google Cloud, and Microsoft Foundry. The company said access would initially be subject to temporary usage limits and later billed through usage credits while it deployed updated safety controls.
The U.S. Commerce Department lifted the export controls previously imposed on Anthropic’s Fable 5 and Mythos 5 models, allowing Anthropic to restore access to Fable. According to the report, Anthropic agreed to detect and address security risks, work with the government on protocols for current and future models, and report malicious activity.
Tom's Hardware reported that Chinese firm Z.ai released its open-weight GLM-5.2 model immediately after U.S. export controls forced Anthropic to disable Fable 5 and Mythos 5 for foreign nationals. The article said GLM-5.2 quickly became a leading legally accessible model for many users outside the United States and that Z.ai claimed it was trained entirely on Huawei Ascend chips.
The U.S. Department of Commerce withdrew export control restrictions on Anthropic’s Fable 5 and Mythos 5 models on 2026-06-30 after Anthropic committed to detect and report security risks, coordinate on release protocols, and disclose malicious activity. The withdrawal ended the 18-day disruption, while Commerce reserved the right to reimpose licensing requirements if conditions change or Anthropic fails to comply.
Austrian State Secretary for Digitalisation Alexander Pröll raised with European Commissioner for Technology Henna Virkkunen a proposal for the EU to encourage Anthropic to establish part of its operations in Europe after the U.S. restricted access to two of its models. The proposal was framed as part of a broader European digital-sovereignty push emphasizing legal certainty, market access, capital, and expanded AI infrastructure.
U.S. officials partially reversed earlier restrictions on Anthropic’s top-tier AI systems, allowing limited access to Claude Mythos 5 for approved American institutions operating essential infrastructure. Anthropic said more than 100 entities would regain access in phases under new guidelines, while unapproved organizations remained subject to licensing rules.
Anthropic said the U.S. government approved restoration of Claude Mythos 5 access for a limited set of U.S. organizations on 2026-06-26. The approval preceded the broader phased return of access later described by officials and the company.
Socket reported that Anthropic may seek a restricted re-release of Fable 5 and Mythos in early July after a 30-day federal review window, using identity verification controls. The report also said the company could introduce a modified 'Mythos 5.1' variant with stricter cybersecurity boundaries.
Nextgov reported that parts of the NSA lost access to Anthropic’s Mythos 5 model amid the administration’s restrictions on Anthropic, affecting at least some red-teaming and cybersecurity testing work. The report said the agency may still retain limited access to earlier model versions under prior arrangements.
French President Emmanuel Macron called on wealthy democracies to cooperate on regulating advanced AI systems and to coordinate on security and cybersecurity issues at a G7-linked meeting. The appeal came amid fallout from the U.S. restrictions on Anthropic’s Fable 5 and Mythos 5 models and reflected growing European concern over dependence on U.S. AI providers.
Wired identified SK Telecom as the South Korean telecom company whose access to Anthropic’s Claude Mythos model was removed at the White House’s request over alleged ties to China. The revocation occurred before the broader export controls that later forced Anthropic to suspend Fable 5 and Mythos 5 for foreign nationals.
After the Defense Department labeled Anthropic a supply chain risk following a dispute over Pentagon access to its models, Anthropic sued the administration. The new reference says that litigation remained ongoing while Commerce Department talks continued over restoring access to Fable 5 and Mythos 5.
Ars Technica reported that OpenAI privately released a cybersecurity-focused model in mid-April 2026 and broadened its cybersecurity strategy. The development was cited by experts arguing that restricting Anthropic alone would not stop similar AI cyber capabilities from emerging elsewhere.
Following the U.S. export restrictions on Anthropic’s Fable 5 and Mythos 5 models, officials in Europe, Canada, and the UK publicly warned that sudden loss of access to frontier AI created a strategic vulnerability for allied governments and projects. The reactions framed frontier AI access as critical infrastructure and highlighted concerns about dependence on U.S. providers.
Lawmakers including Sen. Angus King, Rep. Bennie Thompson, Sen. Mark Warner, and House Homeland Security Chairman Andrew Garbarino publicly reacted to the Trump administration’s export controls on Anthropic’s Fable 5 and Mythos 5 models, with several expressing skepticism about the lack of clear justification and the appearance of a politically driven process. Some Republicans withheld judgment pending more information from the administration.
More than 100 cybersecurity executives and experts urged the Trump administration to reverse or ease export-control restrictions blocking foreign nationals from using Anthropic’s Fable 5 and Mythos 5 models. They argued the limits could weaken U.S. cyber defense more than they impede adversaries and said Anthropic’s models are not uniquely capable compared with other frontier systems.
A U.S. official confirmed the Commerce Department issued the directive targeting Anthropic's Fable 5 and Mythos 5 models, and Pentagon CIO Kirsten Davies publicly supported the action on national security grounds. Her statement added visible Defense Department backing to the export-control dispute.
Axios reported that Amazon raised concerns about the Anthropic export-control dispute to Treasury Secretary Scott Bessent. The report also said follow-up meetings were scheduled with the Commerce Department, the CIA, and White House science adviser Michael Kratsios as the administration and company sought to resolve the standoff.
A coalition of 76 cybersecurity experts published an open letter urging the U.S. government to rescind the export-control order on Anthropic’s Fable and Mythos models. The signatories argued the restriction harms defenders by denying them advanced tools for vulnerability discovery and remediation and called for transparent, scientifically grounded AI regulation.
Anthropic said it complied with the order but challenged the government's rationale, arguing the cited jailbreak was narrow, previously known, and reproducible on other public models such as OpenAI's GPT-5.5. The company said it is working to restore access and warned that applying this standard broadly could disrupt frontier AI deployments across the industry.
Anthropic met with White House officials in Washington, DC, on 2026-06-15 to discuss the export controls on Fable 5, but the talks ended without the Trump administration lifting the restrictions. The outcome showed continued disagreement over the model's jailbreak risk and national security implications.
After receiving the directive, Anthropic globally disabled access to its cybersecurity-focused Fable 5 and Mythos 5 models. ZDNET reported, citing Axios, that Anthropic had about 90 minutes on Friday to take the models offline.
The U.S. government sent Anthropic an unpublished export-control directive invoking national security authorities to bar foreign nationals, including some people inside the United States and Anthropic employees, from accessing Fable 5 and Mythos 5. Anthropic said the government believed the models could be jailbroken for cyber use, but provided only verbal evidence.
Shortly before the export-control action, CEO Dario Amodei published a policy essay supporting government authority to block unsafe AI deployments, though Anthropic later argued such power should be exercised transparently and on a technical basis.
In February 2026, Defense Secretary Pete Hegseth designated Anthropic a supply chain risk following failed military contract negotiations, according to The Record.
Senior technical staff from Anthropic traveled to Washington, D.C. to meet White House officials over the dispute that led to the Fable and Mythos models being taken offline. Axios also reported the company had been holding virtual meetings with administration officials since the administration’s initial outreach on Friday as both sides sought a quick resolution.
According to the new report, Trump had already directed federal agencies in February 2026 to stop using Anthropic’s AI. The move was described as part of broader tensions between Anthropic and the administration before the later export-control action targeting Fable 5 and Mythos 5.
Anthropic said it received a June 12 letter from the U.S. Commerce Department ordering it to suspend access to its Fable 5 and Mythos 5 models for all foreign nationals. According to the report, this effectively triggered a global shutdown because Anthropic could not reliably separate users by nationality.
According to the new report, Anthropic’s Mythos model penetrated nearly all NSA classified systems within hours during an authorized red-team evaluation on 2026-06-11. The alleged result was cited as a key factor behind the U.S. Commerce Department’s subsequent move against Fable 5 and Mythos 5.
Tom's Hardware reported that Anthropic launched its cybersecurity-focused Fable 5 model on 2026-06-10. The launch preceded the U.S. government's export-control actions that followed days later.
Vulnerabilities, threat actors, malware, products, organizations, and breaches Mallory has linked to this story.
50 references tracked. Mallory keeps watching after this page renders.
securityweek.com
Open sourceitpro.com
Open sourcegovconwire.com
Open sourcexakep.ru
Open sourceghacks.net
Open sourcewired.com
Open sourcecio.com
Open sourcecyberscoop.com
Open sourceMap indicators from this story to your assets and identify affected systems in minutes.
Every observed campaign, victim, and pivot linked to actors named in this story.
Malware, exploits, and IOCs connected to the activity described here.
YARA, Sigma, and Snort rules deployed to your SIEM as soon as they’re published.
Get matching new stories delivered to your team as they break — not the next morning.
Ask questions about this story and take action on the answers.
Other stories Mallory is tracking that overlap on entities, sources, or topic labels.
Anthropic restored global access to its **Fable 5** model after temporarily restricting it in response to US government concerns that a jailbreak could enable misuse of the model’s cybersecurity capabilities. The company said Amazon researchers identified a safeguard bypass, prompting Anthropic to redeploy Fable 5 with new classifiers designed to block more cyber-related tasks while reducing false positives. Anthropic also resumed access to **Mythos 5** for a limited set of US organizations and said testing found other, less capable models could identify the same vulnerabilities, suggesting the bypass did not unlock uniquely advanced offensive capability. The episode has intensified scrutiny of how governments and AI vendors assess cyber risk from frontier models. Orion Policy Institute argued that Washington’s temporary ban and reversal exposed a broader weakness in real-world cyber-risk measurement, especially when policymakers must judge whether a model meaningfully increases offensive capability or merely replicates what existing tools can already do. Anthropic said it is now working with the US government and partners including **Amazon, Microsoft, Google, and Glasswing participants** to build a framework for evaluating AI jailbreak severity and determining appropriate response measures.
Jul 11, 2026
Senior U.S. officials are considering giving the Cybersecurity and Infrastructure Security Agency access to Anthropic’s **Mythos** model so it can coordinate vulnerability scanning across federal civilian agencies, with a focus on finding public-facing weaknesses and other security flaws on government networks. The discussions come as agencies seek clearer direction on how advanced AI models will be used in cyber defense, following a White House push for stronger network security measures and CISA’s role in developing guidance and an AI vulnerability clearinghouse. CISA officials have stressed that AI could accelerate detection and prioritization, but would not replace remediation work, infrastructure modernization, or the reduction of technical debt. At the same time, the Trump administration has tightened control over Anthropic’s most advanced systems by placing **Mythos 5** and **Fable 5** under Commerce Department export restrictions, requiring licenses for export, re-export, or transfer to foreign persons. The move reportedly followed national security concerns after claims that Mythos had been jailbroken, and Anthropic said it suspended customer access to the affected models to comply while disputing the government’s rationale. The parallel actions highlight a split U.S. approach to frontier AI: expanding potential domestic cybersecurity use inside government while restricting broader access to the company’s most powerful models.
Jul 12, 2026
The Trump administration has shifted from a largely permissive approach to frontier AI toward tighter scrutiny of models seen as capable of accelerating cyber operations, after private-sector threat intelligence and Anthropic’s release of the cybersecurity-focused **Mythos** raised concerns about risks to critical U.S. infrastructure. According to the reports, officials imposed export controls on Anthropic’s **Fable 5** and **Mythos 5**, explored stricter oversight, and at one point briefly blocked worldwide access to Mythos and OpenAI’s latest model before later lifting those restrictions and moving toward a voluntary pre-release review process. Officials, researchers, and practitioners said advanced models are already helping with vulnerability discovery, code triage, and defensive automation, while also potentially increasing the speed and scale of offensive cyber activity. The coverage says policymakers remain divided over how much these controls can slow comparable capabilities already appearing in open-source and foreign models, and warns that future jumps in AI capability could force harder decisions around critical infrastructure, biological misuse, evaluation standards, government technical expertise, and shortages of advanced compute.
Jul 22, 2026