Vulnerabilities · 80 days ago
OpenAI is pushing AI security work past bug finding and into patch generation, which moves the choke point from discovery to human review. The real limit is no longer how many flaws a tool can surface. It is how many machine-suggested fixes a team can validate, deduplicate, and safely merge before they touch production.
OpenAI released GPT-5.5-Cyber, updated Codex Security, and launched Patch the Planet with Trail of Bits to help trusted defenders and open-source maintainers validate issues and draft patches at scale. OpenAI says Codex Security has scanned more than 30 million commits across 30,000 codebases, with more than 500,000 findings resolved automatically and tens of thousands of fixes confirmed by humans.
For teams already using AI-assisted scanning or patching, the operational risk is a growing review queue, not a new exploit class. Faster intake can outpace the capacity to inspect fixes, and that makes quality control the scarce resource.
8 sources covering this story
Software, AI companies form alliance to tackle open-source security flaws
The emergence of frontier AI models has increased the speed and capabilities of malicious hackers.
Sponsored: Trail of Bits and OpenAI patch the planet
In this sponsored interview James Wilson chats with Trail of Bits founder and CEO Dan Guido about its newly announced partnership with Ope [Read More
OpenAI Expands Daybreak to Help Defenders Patch Flaws
OpenAI expanded Daybreak with a full GPT-5.5-Cyber release to help defenders patch software flaws
OpenAI Refocuses Cybersecurity Efforts on Patching Over Discovery
OpenAI has expanded its Daybreak cybersecurity initiative with a new suite of tools and partnerships.
OpenAI Expands Daybreak With GPT-5.5-Cyber to Help Defenders Patch Security Flaws
OpenAI is releasing GPT-5.5-Cyber, Codex Security updates, and Patch the Planet to help defenders validate and patch vulnerabilities.
OpenAI wants AI to fix vulnerabilities, not just find them - Help Net Security
OpenAI expanded Daybreak with Codex Security, GPT-5.5-Cyber, and Patch the Planet to help find and fix vulnerabilities.
Patch the Planet is our joint initiative with OpenAI to assist critical open-source software in the age of machine-speed vulnerability research.
OpenAI Lets Cyber Vendors Embed GPT-5.5 in Defenses
OpenAI's new Daybreak Cyber Partner Program allows 29 cybersecurity vendors, service providers and integrators to embed GPT-5.5 capabilities into customer-facing
Part of the PlainSec briefing for 2026-06-27