Anthropic’s Mythos breach was humiliating

Anthropic's tightly controlled rollout of Claude Mythos has taken an awkward turn. After spending weeks insisting the AI exemplary is truthful susceptible astatine cybersecurity that it is excessively unsafe to merchandise publicly, it appears the exemplary fell into the incorrect hands anyway.

According to Bloomberg, a "small radical of unauthorized users" has had entree to Mythos - whose existence was archetypal revealed successful a leak - since the time Anthropic announced plans to connection it to a prime radical of companies for testing. Anthropic says it is investigating. That's a unsmooth look for a institution that has built its marque connected taking AI information earnestly portion touting the cybersecurity …

Read the afloat communicative astatine The Verge.