Chainguard broke a coordinated disclosure embargo on PHP flaws. The flaws were found using AI tools in an OpenAI-led process. Cyber agencies in Europe, the US, the UK, Australia and Canada issued urgent warnings. Governments took websites offline as a precaution. Cloudflare saw millions of exploitation attempts within hours. Amazon reported broad exploitation in the wild. OpenAI said more significant vulnerabilities are coming. AI both finds flaws and speeds up their exploitation. Agencies now recommend immediate patching. The incident exposed gaps in global cybersecurity coordination.
What changed
Coordinated disclosure embargoes held until fixes were ready.
What it unlocks
Hackers could exploit PHP flaws at scale using AI tools.
- 70% of the web uses PHP
- millions of exploitation attempts in hours
Sources