OpenAI Expands Daybreak With GPT-5.5-Cyber to Help Defenders Patch Security Flaws

Summary: OpenAI released an improved GPT-5.5-Cyber model to trusted defenders under its Daybreak initiative, saying it can sustain deeper analysis across large codebases to find, validate, and patch vulnerabilities, alongside an updated Codex Security plugin for scanning, triage, and patch generation. OpenAI also launched Patch the Planet with Trail of Bits to secure open-source projects, with initial participants including cURL, NATS Server, pyca/cryptography, Sigstore, aiohttp, the Go project, freenginx, Python, and python.org.
What you should know: The bottleneck has shifted from finding vulnerabilities to patching them, and a 29-year-old flaw in the Squid web proxy (CVE-2026-47729, Squidbleed) shows how long bugs can sit unpatched. Five national cyber agencies warn the timeline for frontier models transforming offensive and defensive capability is months, not years.

Read the original at The Hacker News ↗

Topics: AI Security
Mentions: CVE-2026-47729
Related: 29-Year-Old Squid Proxy Bug 'Squidbleed' Can Leak Cleartext HTTP Requests Cyber Risks ·
⚡ Weekly Recap: Linux Kernel Flaws, AI Malware Tricks, Turla Backdoor, Infostealers and More Cyber Risks ·
SECURITY AFFAIRS AI-CYBERSECURITY NEWSLETTER ROUND 8 Cyber Risks · Nov 15, 2026