OpenAI AI Tools Expose Critical PHP Flaws Affecting Most of the Web

The PHP Foundation moved up the release of critical security patches to October 21 after premature disclosures triggered public concern on October 20. The updated versions, specifically PHP 8.2.35, 8.3.36, 8.4.27, and 8.5.12, address two severe vulnerabilities that could affect an estimated 70% of all websites. OpenAI announced the discovery in a press release on October 21, stating that its GPT-5.6 Sol model helped uncover the flaws as part of its Patch the Planet cybersecurity initiative.

The core issue involves memory management errors within the PHP engine. One flaw occurs when the system runs out of allocated memory, potentially leaving behind incomplete data that the engine later treats as valid. The second involves a fatal error that causes PHP to point to memory that has already been freed. Under specific conditions, attackers can exploit these weaknesses remotely without login credentials to execute malicious code and potentially take control of servers. Both vulnerabilities have been assigned critical severity scores of 9.2 out of 10.

Researchers note that these bugs were not immediately obvious to human developers. The PHP Foundation had been aware of the underlying flaw since 2024 but could not identify a method to exploit it or a way to fix it. It persisted for two years until researchers from Calif, a founding partner in the Patch the Planet program, began auditing the code with OpenAI tools on July 30. Ryan A., a researcher at Calif, first identified the exploitable vulnerability on August 11. OpenAI researchers confirmed the finding on the same day, and a separate firm named Wiz independently identified the second vulnerability shortly after.

The discovery highlights a concept OpenAI calls the Defender's Window. The company argues that there is a narrow timeframe where organizations can use advanced AI models to find and patch defenses before those same capabilities become widely available to malicious actors. OpenAI considers this incident one of the major cybersecurity events of 2026 due to the ubiquity of PHP. The company suggests that while human teams struggled to resolve the code issues for two years, AI-assisted researchers demonstrated practical exploits within just two weeks.

The implications extend beyond simple code errors. Because the flaws exist in the core engine, websites can be vulnerable even if their application code is secure. Researchers demonstrated that the bugs could expose sensitive data stored in server memory, such as passwords, authentication tokens, and API keys. OpenAI maintains that it is a collective responsibility to address such flaws proactively, noting that powerful technology will inevitably be used by bad actors. This follows a recent report that Anthropic's Claude Mythos identified thousands of zero-day vulnerabilities in major operating systems and browsers.

The timeline for the fix was complex, requiring major changes to the PHP engine. OpenAI coordinated a coalition of researchers, maintainers, security providers, and government agencies to address the issues over a two-month period. The company attributed the duration to the technical complexity of the task. Although the coordinated release was originally scheduled for October 22, the early leaks forced the Foundation to act quickly. PHP 8.1 and older versions will not receive official fixes, meaning users on those versions must upgrade to newer releases to stay secure.

OpenAI states that every site using PHP can now safely upgrade to the patched versions. The incident serves as a case study for how AI is transforming the landscape of software security. While some might argue that the flaws would have remained harmless without AI to exploit them, OpenAI challenges this view. The company believes that acting now is essential to strengthen defenses before the gap between defenders and attackers widens further.