✓ VERIFIED — Tom Brown's exact title at Anthropic - source only describes him as 'chief compute officer' without independent verification Multiple independent sources including Bloomberg, Crunchbase, X, and recent reporting confirm Tom Brown is co-founder and Chief Compute Officer at Anthropic. (Bloomberg, Crunchbase, X/Twitter, TechCrunch, Semafor)
— REMOVED — The specific date 'June 12' - verify this is the correct date Anthropic began working with the government Claim was disproven by web check and removed from the article. June 12, 2026 was the date the US government issued an export control directive suspending Mythos 5 and Fable 5, not when Anthropic began working with the government. Tom Brown began Commerce Department talks on June 15, 2026. (CybersecurityNews, AIWeekly, TechCrunch, Semafor, The Conversation)
✓ VERIFIED — The 'more than 100' figure for agencies/companies - confirm this exact count Multiple recent sources confirm the Trump administration cleared Anthropic's Mythos 5 for use by 'more than 100' US government agencies and companies after June 26, 2026. (TechCrunch, Semafor, CNBC, BitcoinWorld)
Source material is straightforward reporting with clear attribution. One minor concern: Tom Brown's role is described but not independently verified; all other facts are directly supported by source text.
tech · lumee
Trump administration allows Anthropic Mythos 5 access for 100+ U.S. entities
Two weeks after pulling advanced AI models from the market, Anthropic gained permission to release Mythos 5 to specific U.S. government agencies and companies under new restrictions.
The short version
Anthropic's Mythos 5 and Fable 5 models were withdrawn after security researchers reportedly bypassed safeguards
The Trump administration authorized Mythos 5 access for more than 100 U.S. government agencies and companies
Non-American employees at authorized organizations can now access Mythos 5, reversing the original ban
Fable 5 remains unavailable; Anthropic continues negotiations for broader access
100+
U.S. government agencies and companies authorized to access Mythos 5
2 weeks
Time between model withdrawal and Trump administration authorization
What happened
Anthropic pulled its Mythos 5 and Fable 5 AI models from the market after security researchers allegedly bypassed their safety guardrails. Two weeks later, the Trump administration authorized Anthropic to release Mythos 5 to more than 100 specific U.S. government agencies and companies, including access for non-American employees at those organizations. The directive, communicated by Commerce Secretary Howard Lutnick, did not address Fable 5's availability.
Context
Both models were withdrawn following reported security vulnerabilities. Anthropic has stated it has been working with the U.S. government to restore access and improve the models' safety measures. The company is continuing negotiations to expand Mythos 5 access further and eventually release Fable 5 for general use.
Both sides
Bull
Structured access allows the U.S. government and vetted organizations to use Anthropic's strongest cybersecurity model while researchers address safety concerns.
Bear
The ban's reversal for selected entities raises questions about whether the underlying security issues have been fully resolved for broader availability.
In plain words
Anthropic had to pause two of its advanced AI tools because hackers found ways around their safety features. The Trump administration now says trusted U.S. government agencies and companies can use one of those tools again, but only for now—the other tool is still locked away while fixes continue.
What may happen
Expanded access granted
Anthropic successfully completes negotiations with the government, resulting in Mythos 5 becoming available to a wider range of U.S. entities and Fable 5 being released for general commercial use.
scenario - not a prediction
Restricted access maintained
Access remains limited to the current 100+ authorized entities for an extended period while the government and Anthropic conduct additional security reviews.
scenario - not a prediction
Renewed security concerns
New vulnerabilities are discovered among authorized users, prompting the government to implement additional restrictions or require further safeguard improvements.
scenario - not a prediction
⚠ Disputed / unconfirmed:
Tom Brown is Anthropic's chief compute officer — The source identifies him only as 'chief compute officer' but does not state his full name or confirm this is his actual title at Anthropic. This is a high-risk proper name/title combination that should be verified independently.
Source: techcrunch.com · original
Generated 2026-06-27 10:02 · lumee