OpenAI just announced a live cyber keynote for November 12. Here's what you need to know. The event is called Cyber at the Frontier. It streams Thursday, November 12 at 10 a.m. PT, with registration on openai.com. OpenAI says it will share its latest cyber announcements there. The pitch is lessons from defending itself. Working with its own frontier models, OpenAI says it found and fixed important vulnerabilities in its own code and infrastructure. It is now turning that experience into tools and playbooks to help other security teams strengthen their defenses. Greg Brockman (@gdb) will share what OpenAI learned. His message is that security teams should keep putting the latest models to work as their capabilities improve. OpenAI has not named a specific product yet. The post went up at 12:10 AM on Nov 11 and passed 254K views. The backdrop is a run of OpenAI security work. It recently introduced GPT-6 Cyber, a model for finding exploitable vulnerabilities and developing fixes. GPT-6 Astra launched September 3, and Sol and Luna followed on September 22. One unverified claim is also circulating. Posts say SoftBank scanned about 700 systems and found 10,500 vulnerabilities, held an emergency press conference on June 16, and spent three months patching. They say a re-scan with GPT-6 Astra then found about 35,000 more. The reaction is mostly about models. Many users are asking where GPT-6.2 Astra and Sol are, since GPT-6.2 appears in OpenAI's docs. Others note Anthropic's Fable 5.2 and Google's Gemini 4 Carbon are already out. Key numbers: - Keynote: November 12, 10 a.m. PT - Original post: 254.3K views - Astra launch: September 3, 2026 - Sol and Luna launch: September 22, 2026 - Claimed SoftBank scan: 700 systems, 10,500 vulnerabilities OpenAI's cyber announcements land Thursday, and no new model has been confirmed for the event.