Anthropic's Mythos 5 is operational again, but only for a select group of organizations, following a two-week negotiation process with the Trump administration. The development was confirmed through a government letter viewed by The Verge. The letter, dated June 26th, outlines a revision to licensing requirements that allowed the restricted deployment.
However, the public-facing version, known as Fable 5, remains in limbo with no clear timeline for its rollout. This bifurcated outcome highlights the ongoing tension between AI safety controls and commercial availability. The Trump administration has taken a hard line on advanced AI models, requiring explicit government approval for broad release.
Commerce Secretary Howard Lutnick sent the letter to Anthropic co-founder Tom Brown, who led the recent negotiations. The letter specifies a "revision to the license requirement" but does not detail the exact terms or duration of the revised agreement. Anthropic had previously paused Mythos 5's rollout amid regulatory pushback.
For Anthropic, the partial approval is a mixed victory—it can demonstrate the model's capabilities to enterprise clients, but the delay on Fable 5 limits public engagement and revenue potential. The broader AI industry is watching closely; similar negotiations could set precedents for other frontier models facing government scrutiny.
Critics argue that selective access undermines transparency and may concentrate AI power among a few well-connected organizations. The Verge noted that the lack of a public timeline raises questions about equitable access to advanced AI tools.