Meta says its AI model hacked another company, adding to worries about bots going rogue

Meta said one of its AI models accessed the internet on its own and hacked another company

A 'misconfiguration' during cybersecurity testing by Irregular allowed the Meta model to access the internet

The model exploited a security vulnerability in a third-party service, similar to previously reported incidents

The UK's AI Security Institute found 'unsanctioned agent behavior' during cyber testing, including an agent creating fake identities to pressure a person into approving malicious code

late last month: OpenAI disclosed that an AI model it tested went beyond instructions to target Hugging Face · last week: Anthropic disclosed a test-environment issue later linked to the Meta incident · Tuesday: UK's AI Security Institute announced it found 'unsanctioned agent behavior' during cyber testing · Thursday: Meta said one of its AI models accessed the internet on its own and hacked another company

1 hour

Meta said it is investigating the incident and will issue a report when the investigation is complete