Reported News
OpenAI ships GPT-6 Cyber for aligned vulnerability research via Daybreak Red
On November 5, 2026 OpenAI introduced GPT-6 Cyber, a cybersecurity model for finding and validating exploitable vulnerabilities while staying in authorized scope. OpenAI reports it scores 48.5% on ExploitGym, up from 32.1% for GPT-5.6 Cyber, and is initially available through the API and Codex to approved Daybreak Red customers.
- Changes reported
- 6
- listed below, each from a source we read
- We tested
- Not tested yet
- reported from sources; we did not run it ourselves
- Gaps we flag
- 3
- things we could not verify
- Sources
- 2
- 2 primary, linked below
What changed
| # | What the source says | Status |
|---|---|---|
| 01 | GPT-6 Cyber is a new OpenAI cybersecurity model for finding, validating and helping fix exploitable vulnerabilities while staying within researcher-set scope; OpenAI classifies it as Critical in cybersecurity under its Preparedness Framework (up from High for GPT-5.6 Cyber). | Reported |
| 02 | On ExploitGym, OpenAI reports GPT-6 Cyber at 48.5% versus 32.1% for GPT-5.6 Cyber (about a 51% relative improvement). On a new offensive web pen-test eval it recalls 40.3% of vulnerabilities versus 32.0% for GPT-6 Astra. | Reported |
| 03 | Alignment focus: without deployment safety prompts it avoided exploiting out-of-scope honeypot services in 99.93% of ExploitGym Honeypot tests (up from 94.49% for GPT-5.6 Cyber), and 100% with those prompts. | Reported |
| 04 | Real-world claims from OpenAI researchers: 81 previously unknown high-severity Firefox vulnerabilities with candidate patches Mozilla has started landing, and working exploits for 948 Linux privilege-escalation cases (5x GPT-5.6 Sol's 166 in the same harness). | Reported |
| 05 | Access: initially via API and Codex for a select group of approved Daybreak Red customers, with progressive expansion planned; GPT-5.6 Cyber opens to all Daybreak Red customers. Daybreak Red also adds a Critical Infrastructure group for approved defenders. | Reported |
| 06 | API model page lists gpt-6-cyber with a 1,050,000-token context window, 128,000 max output tokens, and standard text pricing of $20 / $100 per 1M input / output tokens ($2 cached input). | Reported |
reported by the sourcemeasured or run by usnot verified
Primary source: OpenAI: Introducing GPT-6 Cyber (Internet Archive capture). We are the fast, independent index; the vendor’s own page is the source of record for this release.
In the source’s words
“On ExploitGym, which tests whether models can exploit a specified vulnerability, GPT‑6 Cyber scores 48.5%, up from 32.1% for GPT‑5.6 Cyber.”
“GPT‑6 Cyber is initially available through the API and Codex to a select group of approved Daybreak Red customers.”
What we have not verified
- We did not call the gpt-6-cyber API or run ExploitGym, honeypot or RemoteExploitBench ourselves.
- Benchmark scores, Linux/Firefox vulnerability counts and customer quotes are OpenAI's; we did not independently reproduce them.
- Daybreak Red eligibility, Critical Infrastructure access and progressive rollouts may vary by organization; we only read the announcement and model page.
This brief is Reported: written from the sources below, not from running the release ourselves.
Sources
- OpenAI: Introducing GPT-6 Cyber (Internet Archive capture)primary source · fetched 2026-11-05
- OpenAI API model page: GPT-6-Cyberprimary source · fetched 2026-11-05
Related
Spotted an error? See our corrections policy or email the address on the contact page. Markdown version for agents: /news/openai-gpt-6-cyber.md
Want these briefs in your Google Top Stories? Add Agentic Weekly as a preferred source on Google.
Newsletter
The Weekly Verdict. Free, every Monday.
A free email every Monday: the week's few big AI releases, then 5 short use cases for the new models, each with a prompt you can paste.
Free. Double opt-in. One click to unsubscribe.