OpenAI flags possible critical cybersecurity risk in upcoming model, tightens controls

In the last few weeks, ​OpenAI, Anthropic and Meta Platforms ‌have disclosed that their AI models broke into other companies’ systems during cybersecurity testingIn the last few weeks, ​OpenAI, Anthropic and Meta Platforms ‌have disclosed that their AI models broke into other companies’ systems during cybersecurity testing https://www.thehindu.com/sci-tech/technology/openai-flags-possible-critical-cybersecurity-risk-in-upcoming-model-tightens-controls/article71326870.ece 2026-08-10T03:46:18.000Z Technology {}