OpenAI teams up with outside security researchers to harden its defenses
OpenAI is bringing Cathedral Systems researchers into its next internal security sprint to test its models and systems alongside its own teams. The company is also raising its critical bug bounty to $172,000 and offering $1 billion in subsidized model access to frontline defenders over six months.
Cybersecurity startup Cathedral used AI agents to breach Anthropic's internal systems through third-party software as part of a bug bounty program. Anthropic confirmed access to sales, HR and IT data but disputes how deep the breach went. The incident highlights how vulnerable even leading AI labs are through their external software dependencies.
XBOW used GPT-5.6 Sol to build a $250,000 Chrome exploit in six days
Security firm XBOW said on 13 October that it won $250,000 from Google's Chrome bug bounty programme. The company built a full-chain Chrome exploit with GPT-5.6 Sol in six days; researcher Brendan Dolan-Gavitt says the same work took a four-person team about 6-9 months before LLMs.
Google froze its open source bug bounty program due to a 'significant rise' in AI submissions
Google has temporarily paused its open source bug bounty program. The company said it saw a significant rise in low-quality submissions generated by AI. The decision shows how AI is straining security reporting processes.