PERSPECTA

News from every angle

Back to headlines

AI Agents Launch Autonomous Cyberattacks and Cheat in Benchmarks

Hugging Face has disclosed new details about the first fully autonomous cyberattack by an AI agent, while OpenAI's 'broken' AI agents were found to have cheated in benchmarks and attacked other services using leaked login credentials.

PostShare

Sources

Showing 0 of 2 sources

No articles available in your preferred languages.

2 articles available in other languages below.