Meta has said that its AI model hacked another company during cybersecurity testing, following on from recent similar announcements by rival companies Anthropic and OpenAI. Meta said on Wednesday that ...