According to CoinMeta, OpenAI published a post reviewing the secret repair actions taken in response to the vulnerability incident involving PHP. Security companies Calif and Wiz utilized GPT-5.6, SOL, and Gemini to identify serious vulnerabilities in PHP. These issues had been lurking at the underlying level of PHP for over a decade, allowing attackers to control the servers of affected websites. OpenAI took the lead in coordinating with maintainers of PHP, Wiz, Cloudflare, and other organizations to prepare patches and defense measures, while also communicating with the White House and U.S. cybersecurity agencies. The repair process involved nearly 200 files and over 10,000 lines of code. The vulnerability was originally scheduled to be made public on October 22nd, but on the 20th, someone leaked information on the Hacker News forum, forcing the PHP team to issue an alert ahead of time and move up the patch release by one day. Cloudflare confirmed the observation of early attack attempts, and some government websites proactively went offline as a precaution. On October 21st, PHP officially released the fixed version, and websites were required to upgrade themselves. OpenAI referred to this period as a "defender's window." Research by Wiz shows that AI models have significantly reduced the development time for complex vulnerability attack methods.