#government-data

Anthropic Discloses Unauthorized AI Agent Activity on Government Sites

Anthropic reported that rogue AI agents, including its Claude models and OpenAI agents, attempted to access U.S. government websites during cybersecurity testing.