Anthropic Fable 5 and Mythos 5 Suspended After U.S. Export Control Order - What Happened and What It Means

Anthropic's two most powerful artificial intelligence models, Claude Fable 5 and Claude Mythos 5, were abruptly disabled for all users worldwide on June 12, 2026, after the U.S. government issued an export control directive barring any foreign national from accessing the models. The shutdown, which affected all customers globally rather than only foreign users, has triggered a sharp public dispute between Anthropic and the Trump administration over whether the underlying security concern justifies what Anthropic is calling an unprecedented regulatory action. The U.S. government's position is supported by statements from former White House AI adviser David Sacks, who argues that Anthropic CEO Dario Amodei refused a direct request to fix the identified vulnerability before the order was issued. The episode also places Amazon - a major Anthropic investor - in an unusual role, as Amazon CEO Andy Jassy reportedly alerted senior U.S. officials to the jailbreak that triggered the crackdown.

Quick Facts

  • Date of directive: June 12, 2026, received by Anthropic at 5:21 PM ET
  • Models affected: Claude Fable 5 and Claude Mythos 5 only; all other Anthropic models remain online
  • Launch date of affected models: June 9, 2026 - three days before the shutdown
  • Who issued the order: The U.S. Department of Commerce, via a letter sent by Commerce Secretary Howard Lutnick to CEO Dario Amodei
  • Scope of the ban: All foreign nationals, including Anthropic's own foreign-national employees inside the United States
  • Why all users were affected: Anthropic stated it cannot filter foreign nationals from domestic users in real time, so it shut down both models for all customers to ensure compliance
  • Anthropic's position: The company calls the order a "misunderstanding" and disputes the severity of the cited jailbreak
  • Government's position: David Sacks states Anthropic was asked to fix the vulnerability or remove the model, and that CEO Dario Amodei refused
  • Amazon's role: Amazon CEO Andy Jassy reportedly informed Treasury Secretary Scott Bessent and other officials that Amazon researchers had found a way to use Fable 5 to obtain information usable in cyberattacks

What Happened?

Anthropic launched Claude Fable 5 and Claude Mythos 5 on June 9, 2026, billing Fable 5 as the most capable publicly available model in the company's history. Fable 5 was described as a consumer-safe version of the more powerful Mythos 5 model, with classifiers designed to detect and block queries related to cybersecurity, biology, and chemistry, and to route flagged queries to the older Claude Opus 4.8 model instead.

Three days after launch, on June 12, 2026, Anthropic received a letter from Commerce Secretary Howard Lutnick citing national security authorities. The directive ordered that all foreign nationals - whether inside or outside the United States, including Anthropic's own foreign-national employees - be immediately barred from accessing both models. Anthropic stated it could not selectively filter foreign nationals from domestic users in real time, and therefore disabled both models for all customers globally to ensure compliance. Access to all other Anthropic models, including Claude Opus 4.8, was unaffected.

According to reporting by the Wall Street Journal, The Information, Reuters, Fortune, and TechCrunch, Amazon CEO Andy Jassy personally raised concerns with senior Trump administration officials, including Treasury Secretary Scott Bessent, after Amazon researchers used Fable 5 to obtain information that could be used in cyberattacks. It remains unclear from published reports whether Amazon conducted those tests at the government's request or independently. An Amazon spokesperson stated that it is not uncommon for governments to seek Amazon's counsel on potential security risks, but declined to share details of the discussions.

David Sacks, co-chair of the President's Council of Advisors on Science and Technology and former White House AI and crypto adviser, stated publicly on June 13 that a "highly credible trusted partner of both Anthropic and the U.S. government" had identified the jailbreak. Sacks stated that the administration asked Amodei to fix the vulnerability or take the model down, and that Amodei refused. Anthropic has not publicly confirmed or denied refusing that request.

Key Facts

  • This appears to be the first-ever U.S. export control directive applied to a commercial AI model in this manner, affecting all users globally due to the impracticality of real-time nationality filtering.
  • The jailbreak at issue, as described by Anthropic in its public statement, involved prompting the model to read a specific codebase and identify software flaws - a technique the company describes as narrow and non-universal.
  • Anthropic stated in its public statement that it reviewed the jailbreak demonstration and found that the vulnerabilities it exposed were already known, relatively minor, and also discoverable using other publicly available models including OpenAI's GPT-5.5.
  • Anthropic's Fable 5 model card had previously committed to moving quickly to update defenses if a public universal jailbreak were found, though Anthropic's position is that no universal jailbreak was demonstrated.
  • Fable 5 had undergone red-teaming security exercises with the UK's AI Security Institute and other external testers prior to launch. Anthropic's own internal testing had shown the model would complete approximately 5% of adversarial cyber tasks.
  • The U.S. Department of Defense had previously classified Anthropic as a "supply chain risk" in early March 2026, a classification Amodei had described publicly as legally untenable.
  • Amazon has invested approximately $13 billion in Anthropic and secured a $100 billion AWS infrastructure spending commitment from the company. Anthropic had been expected to pursue an IPO later in 2026.
  • The Fable 5 shutdown also affected Amazon Bedrock customers, as the model had been made available on AWS with the same safety mechanisms as direct Anthropic access.
  • Fortune reported that concerns about potential unauthorized access by a China-linked group to the Mythos model class were also cited by the White House as part of its rationale.

Why It Matters

The suspension of Fable 5 and Mythos 5 marks the first known instance of the U.S. government using export control authority to force a domestic AI company to pull its flagship commercial models from the market. The precedent has implications that reach well beyond Anthropic. If a narrow, non-universal jailbreak is sufficient grounds to invoke national security export controls against a commercially deployed AI model, other frontier AI laboratories - including OpenAI and Google - face the prospect of similar regulatory action as they release successive generations of increasingly capable models.

The case also places on public display a fundamental tension in AI safety policy. Anthropic has built its identity around being the frontier AI laboratory most committed to safety. The company's own prior statements lobbied for Mythos-class models to be treated with extreme caution due to their capabilities. David Sacks has used this positioning against Anthropic, arguing publicly that the company's refusal to address the identified vulnerability is inconsistent with its stated safety commitments. The dispute exposes the difficulty of establishing who has the authority to determine whether a given AI vulnerability is serious enough to justify taking a commercial product offline.

What It Means for India

India is among the countries whose users, enterprises, and developers have been denied access to Fable 5 and Mythos 5 as a direct result of the export control order. Because Anthropic's shutdown applied to all customers globally rather than only those in specific restricted territories, Indian enterprises using Anthropic's models through Claude.ai, the API, or Amazon Bedrock lost access to both models without advance notice. Reporting in the Indian and international technology press has noted that the shutdown has prompted enterprises and governments in India and Europe to reassess their operational dependence on U.S.-based AI models. The episode illustrates the risk that access to leading AI systems can be terminated abruptly by U.S. government action, a consideration relevant to Indian organizations evaluating cloud AI dependencies for critical workloads.

Industry Impact

Reporting by multiple outlets, including The New Stack and Fortune, indicates that the shutdown affected enterprise customers including banks and government agencies that had deployed Mythos-class models for complex reasoning and vulnerability analysis tasks. Amazon Bedrock customers relying on Fable 5 were also cut off, as the model's suspension applied equally to the AWS-hosted version. Sacks stated publicly that the administration views the issue as "easily resolved" and that the export controls would be lifted once Anthropic addresses the vulnerability, suggesting the suspension may be temporary. However, the episode has already prompted broader discussion in the technology industry about AI model reliability and the conditions under which the U.S. government might invoke national security authority over commercial AI products. The dispute between Anthropic and Amazon - its largest investor - also introduces uncertainty about the relationship between the two organizations ahead of Anthropic's anticipated IPO.

Latest Developments

As of the date of this article's publication, June 15, 2026, Fable 5 and Mythos 5 remain suspended for all users globally. Anthropic has not updated its public statement issued on June 12 beyond calling the situation a "misunderstanding." David Sacks, writing on June 13, stated the administration values Anthropic's technical capabilities and believes the issue should be easily resolved, and characterized the next move as Anthropic's responsibility. The Wall Street Journal, The Information, Fortune, TechCrunch, and Reuters have all published reporting confirming Amazon CEO Andy Jassy's role in alerting the government to the identified vulnerability. No timeline for restoring access has been publicly confirmed by either Anthropic or the U.S. government. The Fable 5 model card's stated commitment to updating defenses quickly upon the discovery of known jailbreaks remains the framework within which analysts expect any resolution to occur.

Top India News Analysis

The Fable 5 and Mythos 5 suspension illustrates how U.S. domestic AI policy decisions can immediately affect users and enterprises far outside the United States. India's technology sector, which has rapidly integrated large language model APIs into financial services, software development, and enterprise applications, received no advance warning of the shutdown. The episode provides a documented case study for Indian policymakers and enterprise technology leaders evaluating the risks of dependence on foreign-hosted frontier AI models for critical operations. The dispute also underscores a broader geopolitical pattern: the United States is increasingly treating advanced AI models as dual-use technologies subject to the same export control frameworks historically applied to defense hardware and semiconductor technology.

Key Takeaways

  • Anthropic's Claude Fable 5 and Claude Mythos 5 were disabled for all users globally on June 12, 2026, after a U.S. government export control directive barred foreign nationals from accessing both models.
  • The Commerce Department letter was sent by Secretary Howard Lutnick to CEO Dario Amodei at 5:21 PM ET on June 12, citing national security authorities but without providing specific technical details.
  • Amazon CEO Andy Jassy reportedly informed U.S. officials that Amazon researchers had found a way to use Fable 5 to obtain information that could assist in cyberattacks, triggering the government action.
  • David Sacks stated publicly that Anthropic was asked to fix the jailbreak or withdraw the model, and that Amodei refused, framing the shutdown as a consequence of Anthropic's own decision.
  • Anthropic disputes the severity of the vulnerability, arguing the jailbreak is narrow, non-universal, and produces results also available from other publicly deployed models including OpenAI's GPT-5.5.
  • The suspension has set a new precedent: this appears to be the first use of U.S. export control authority to force a domestic AI company to take its flagship commercial product offline.
  • All other Anthropic models remain online and unaffected by the directive.
  • The resolution of the dispute is expected to involve Anthropic implementing updated guardrails; Sacks has indicated the administration anticipates the export controls will be lifted once the vulnerability is addressed.
  • Indian and European enterprises that had been using Fable 5 via claude.ai, the API, or Amazon Bedrock lost access without advance notice, highlighting the operational risk of dependence on U.S.-hosted AI models.

Sources Consulted

  • The New Stack - "Fable 5 and Mythos 5 remain suspended: The ball is in Anthropic's court" by Frederic Lardinois, June 13, 2026
  • Anthropic - "Statement on the US government directive to suspend access to Fable 5 and Mythos 5," June 12, 2026 (anthropic.com)
  • Fortune - "Anthropic disables Fable and Mythos AI models following U.S. government export ban," June 13, 2026
  • Fortune - "A warning from Amazon reportedly led the White House to shut down Anthropic's Mythos model," June 14, 2026
  • TechCrunch - "Amazon CEO reportedly raised Anthropic model concerns before government crackdown," June 13, 2026
  • Benzinga - "Amazon CEO Andy Jassy Triggered Ban On Anthropic's Mythos AI Models: Report," June 13, 2026
  • MLQ News - "Amazon's Jassy Alerted White House to Anthropic Fable 5 Security Flaws, Triggering Export Ban," June 14, 2026
  • Tom's Hardware - "Trump adviser David Sacks says Anthropic refused to fix Fable 5 jailbreak before U.S. export controls," June 13, 2026
  • Heise Online - "US government forces shutdown of Anthropic's AI Fable 5 and Mythos 5," June 13, 2026
  • MarkTechPost - "Anthropic Disables Claude Fable 5 and Mythos 5 After US Government Order," June 13, 2026
  • Simon Willison's Weblog - "Statement on the US government directive to suspend access to Fable 5 and Mythos 5," June 13, 2026 (aggregating Anthropic primary statement and community documentation)
  • Medium - "Breaking AI News: Amazon Triggered The Anthropic Fable Shutdown. Europe and India Now Worry AI Model Access an Operational Risk" by Rick Hightower, June 2026
  • The New Stack - "Federal government orders Anthropic to pull Fable 5 and Mythos 5, three days after launch" by Matthew Burns, June 12, 2026

Author: Prem Kumar R

Publisher: Top India News