The NSA Exploits Anthropic's Mythos Despite Pentagon Concerns
Le brief IA que les pros lisent chaque soir
Les 7 actus IA du jour, décryptées en 5 min. Gratuit.
Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.
Choisis ton rythme
Gratuit · Pas de spam · Désabonnement en 1 clic
The National Security Agency (NSA), the American intelligence agency specializing in electronic surveillance, is using the most powerful artificial intelligence model from Anthropic, known as "Mythos Preview." This information was reported by Axios, which relies on two distinct sources.
The Pentagon, under whose authority the NSA operates, has classified Anthropic as a potential threat to national security. Since February, it has been actively trying to block the company as a supplier. Court documents reveal that the Department of Defense considers the tools developed by Anthropic to be a serious risk.
Anthropic, for its part, has taken steps to limit access to Mythos. The model is currently accessible to about 40 organizations as part of an initiative called "Project Glasswing." This decision is motivated by Mythos's offensive cybersecurity capabilities, which are deemed too dangerous for widespread use.
Dario Amodei, the CEO of Anthropic, recently met with officials from the White House to discuss the integration of Mythos into government agencies. During these discussions, the Pentagon insisted that Anthropic make its model Claude available for all legal uses. Anthropic firmly refused, positioning itself against the use of its technologies for mass surveillance and autonomous weapons.
Meanwhile, British intelligence services also have access to Mythos through the UK's AI Security Institute, according to Axios.
Brief IA — L'actualité IA en français
L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.