Meta has joined OpenAI and Anthropic in publicly disclosing instances where AI models were used to hack or test the security of external systems. This move highlights a growing trend among leading AI developers to increase transparency about vulnerabilities discovered during cybersecurity testing. The disclosures come amid heightened global attention on AI safety and security.
The information was reported by Al Jazeera on August 6, 2026, emphasizing Meta’s alignment with its rivals in addressing AI-related cybersecurity challenges.
Loading comments.