Which AI model found the PHP vulnerability?

Security firm Calif used OpenAI's GPT-5.6 Sol model as part of OpenAI's Patch the Planet project.

Did anyone else find the same vulnerability?

Yes. Security firm Wiz independently found it, using GPT-5.6 Sol as well as Google's Gemini 4 Argon.

How long did the fix take to prepare?

OpenAI worked with partner companies and the US government for two months to prepare the fix and its release.

Large Language ModelsAI Coding AssistantsAI Safety & AlignmentITmedia AI＋10m agoOct 22, 2026, 04:00 JST

OpenAI AI finds PHP flaw via Patch the Planet

Send on LINE

3 Key Points

What happened
OpenAI said the PHP vulnerability, fixed the same day, was found using GPT-5.6 Sol through its Patch the Planet project with security firm Calif.
Why it matters
OpenAI says frontier AI got there before attackers, but attackers can use AI too, so it predicts such findings will surge in the coming months.

WHO IT HITSSoftware teams that build on open-source languages like PHP may see a faster flow of disclosed flaws and urgent patch releases, while security researchers and AI vendors gain a new front in vulnerability discovery.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, in your inbox every morning.

Get news like this every morning, free

Context & Analysis

OpenAI framed the PHP discovery as part of Patch the Planet, a project to support security in open-source software, and it worked for two months with partners, including the US government, to line up the fix released the same day. The finding was not unique to OpenAI's tools: security firm Wiz also found the vulnerability independently, using GPT-5.6 Sol and Google's Gemini 4 Argon. OpenAI's own framing is double-edged — frontier AI reached the flaw before attackers, but attackers can also use AI, which underpins its prediction that similar discoveries will surge in the coming months. In response, OpenAI says it will continue support through its security program and work on frameworks for vulnerability response built around AI.

FAQ

Which AI model found the PHP vulnerability?: Security firm Calif used OpenAI's GPT-5.6 Sol model as part of OpenAI's Patch the Planet project.
Did anyone else find the same vulnerability?: Yes. Security firm Wiz independently found it, using GPT-5.6 Sol as well as Google's Gemini 4 Argon.
How long did the fix take to prepare?: OpenAI worked with partner companies and the US government for two months to prepare the fix and its release.

ITmedia AI＋Read Original Article

Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Get it free

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.