Apple’s public security records credit only three WebKit vulnerabilities to OpenAI Codex Security, and the widely circulated claim that 35 vulnerabilities in Apple devices were found has not been confirmed by public records. In November 2026, OpenAI announced that its autonomous cybersecurity process had discovered 948 Linux kernel local privilege escalation vulnerabilities and generated exploits and patches. The related operation took place in October and consumed approximately $1.2 million in API credits.
Source:Crypto Briefing · Read Original
Leave a Reply