OpenAI to host cybersecurity summit with 300 leaders to address ai-driven threats

OpenAI is set to host a cybersecurity summit on September 3, 2026, bringing together approximately 300 security leaders from Fortune 100 companies. The event, scheduled for Thursday at OpenAI’s headquarters, will focus on how organizations can leverage AI tools to enhance their defenses against cyber threats. The summit comes amid growing concerns over AI-powered attacks on critical infrastructure, following recent incidents such as the unexpected breach of Hugging Face and suspected AI-driven attacks on U.S. water systems. OpenAI’s CEO, Greg Brockman, is expected to announce a plan to expand access to its tools for public sector institutions. The initiative follows a warning from OpenAI in a public letter, urging organizations to prepare for AI-enabled cyber threats within months. The summit also coincides with the release of the Astra model, which has achieved 'critical' cybersecurity capabilities. Additionally, OpenAI is exploring new training methods that raise concerns about whether security professionals will still be able to understand model reasoning.