OpenAI model finds two critical PHP flaws, triggering patch rush

The episode stands out less for the flaws themselves than for how they were found. PHP underpins a large share of the world's websites, so a critical flaw in the language puts a wide range of companies and government agencies on alert at once. The use of a powerful OpenAI model to surface the vulnerability is described as one of the most high-profile demonstrations of AI's cybercapabilities.

The response was immediate: organizations rushed to patch the two flaws last week. That speed reflects the stakes for anyone running PHP-based sites, but it also points to a broader shift. As AI tools find software flaws faster than ever, the episode offers an early look at how AI-driven discovery may reshape the pace of security work — a theme Michael Daniel, president of the Cyber Threat Alliance and a onetime White House cybersecurity coordinator, discussed in remarks about what the case means.