PHP: OpenAI Warns of Vulnerability Affecting 70% of Websites

Critical vulnerabilities in PHP, identified using an OpenAI model, lead to shutdowns and security alerts. Early attacks have been reported, while authorities recommend taking systems offline or updating them. OpenAI estimates that up to 70% of websites could be affected.
Websites Shut Down, Signs of Attacks, and Calls to Action Starting Wednesday
A large number of websites, including government portals and media outlets, were disabled on Wednesday to address the vulnerability detected in PHP. Dane Knecht, Chief Technology Officer at Cloudflare, reported that initial attacks had been observed and that mitigation measures were being adjusted in response to the emergence of new vectors. The public disclosure of the vulnerabilities prompted urgent messages from leaders, including Japan's Prime Minister. Sanae Takaichi called for immediate action, stating, "We have an urgent request for everyone to act tonight." This announcement triggered a rapid mobilization to implement updates and patches across many organizations.
Guidance from Agencies and Potential Scope of the Risk
The FBI and NSA have advised website operators to take their platforms offline or update their software for protection. The NSA's cybersecurity division specified that service providers using PHP should either take their systems offline or apply patches, following the identification of two critical vulnerabilities in the language's infrastructure. OpenAI estimates that 70% of global websites could be affected by these vulnerabilities. PHP, an open-source language widely used for managing user connections and online commands, exposes many services to significant risks.
Le brief IA que les pros lisent chaque soir
Les 7 actus IA du jour, décryptées en 5 min. Gratuit.
Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.
Choisissez votre rythme
Gratuit · Pas de spam · Désabonnement en 1 clic
Who Discovered the Vulnerability and with What Tools
The critical vulnerability was identified by Calif, a cybersecurity research firm using OpenAI's GPT-5.6 Sol model. Wiz, another company in the sector, also independently discovered these issues, according to OpenAI. Both companies are participating in the "Patch the Planet" initiative led by OpenAI, which aims to enhance the security of global digital infrastructure.
What OpenAI and Its President Say About the Threat
Greg Brockman, President of OpenAI, indicated that the vulnerability could be exploited by malicious actors to steal private user information or alter websites. OpenAI describes this event as one of the major security incidents in recent years. The organization notes that the vulnerabilities appear to have been identified before attackers discovered them.
Brief IA — L'actualité IA en français
L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.