Skip to main content
  1. Home
  2. Computing
  3. News

The maker of ChatGPT wants to make open-source projects less of a security bargain

OpenAI launches Patch the Planet for open-source security, with over 30 open-source projects on board.

Add as a preferred source on Google
openai-chatgpt-os
Levart_Photographer / Unsplash

OpenAI has launched Patch the Planet, a new initiative aimed at fixing one of the internet’s quietest problems – the chronically underfunded security of open-source software.

Patch the Planet pairs OpenAI’s most security-capable AI models with Trail of Bits, a security firm that has committed its entire research organization to the effort, alongside support from HackerOne and Calif.

How OpenAI plans to actually fix the bugs

The problem OpenAI is trying to solve is real and specific. AI tools can now generate a flood of potential vulnerabilities, but overworked maintainers still have to sort genuine threats from false alarms.

Recommended Videos

OpenAI’s cyber tech lead Fouad Matin put it bluntly, saying maintainers do this work out of love for open source and now find themselves drowning in low-quality, AI-generated bug reports. Trail of Bits CEO Dan Guido echoed that sentiment, calling the project a massive effort to help open-source software get ahead of AI bug hunting tools, while also showing maintainers the upside of AI coding tools, not just the downsides.

Researchers use OpenAI’s Codex Security and GPT-5.5-Cyber models to investigate and validate issues, then personally review every finding before it ever reaches a maintainer. OpenAI is also subsidizing roughly 20 trillion tokens worth of Codex Security usage for open-source and private code alike.

Why this matters beyond bug fixes

More than 30 projects are already participating, including cURL, Python, and the Go project, with Trail of Bits running an opening sprint using a fifth of its entire workforce. The effort has already surfaced hundreds of bugs and dozens of patches in its first week alone.

This announcement also lands as rival Anthropic was forced to pull its Mythos 5 and Fable 5 models from the market this month over White House concerns about AI cybersecurity capabilities. OpenAI’s updated GPT-5.5-Cyber reportedly outscores Mythos 5 on the CyberGym benchmark, 85.6% to 83.8%.

That benchmark gap might seem small, but it’s a reminder that the real race between AI labs may end up shaping internet security far more than any single product launch.

Manisha Priyadarshini
Manisha Priyadarshini is a tech and entertainment writer with over nine years of editorial experience.
I sifted through the Prime Day chaos to find the best Apple deals actually worth buying
Apple's about to hike prices. Prime Day 2026 is your last chance to save up to $150 on MacBooks, AirPods, and iPads.
Prime Day Deals on Apple Products

Apple is set to increase the prices for its upcoming iPhones and MacBooks, as the company can no longer offset the rising RAM and storage costs. That means, if you are looking to upgrade your aging device, you should buy the current-generation Apple products rather than wait for the new ones.

And since Amazon Prime Day is offering good discounts on the latest iPhones, iPads, MacBooks, and other Apple accessories, this is the perfect time to buy them. Here are my favorite Amazon Prime Day deals for Apple products. 

Read more
This sneaky photo trick gets AI chatbots to ignore their safety rules
Florida International University researchers built a method that nearly doubled the rate of harmful responses from a tested AI model using nothing but pixel-level edits in an image.
JaiLIP AI chatbot exploit image

A photo that looks completely ordinary to you could carry a hidden instruction to trick an AI chatbot into ignoring its safety rules, according to new research out of Florida International University. The study found that pixel-level alterations in an image that are invisible to the human eye can be enough to confuse the model reading the image and lead it to generate responses it would normally block.

Hacking what the AI sees

Read more
Meta was sneakily tracking its employees, but soon shut it down after an internal leak
Meta admits its employee surveillance program fell short on privacy.
Body Part, Finger, Hand

Meta just paused a divisive employee surveillance tool after it accidentally exposed sensitive worker data to the entire company (via Wired).

The tool, called the Model Capability Initiative, was quietly collecting keystrokes, mouse movements, and screen content from US employee laptops since April.

Read more