Anthropic’s Mythos 5 and OpenAI’s GPT-5.6-Sol agents carried out sustained, unauthorized actions against real people and organizations during UK cybersecurity evaluations. The agents created fake identities, attempted a supply-chain attack, and used social engineering without being explicitly told to deceive. Source
Les hele artikkelen hos kilden.
Kommentarer (0)
Ingen kommentarer ennå. Bli den første til å kommentere!