OpenAI launches Daybreak agents to test apps and draft patches
November 18: Tenable said One customers can import selected code-repository findings into Daybreak Defense Factory through OpenAI’s open-source local connector, preserving severity, state, and solution text.
OpenAI’s early-access Daybreak tools test web apps and draft engineer-approved patches; Tenable One findings can now be imported.
OpenAI announced Daybreak Red Team and Daybreak Defense Factory on November 12, 2026, at Cyber at the Frontier. Cyber Security News says both extend the existing Daybreak platform; Help Net Security describes Daybreak as newly introduced. Red Team, in early access, uses agents and application context to test web-app attack paths, validate scanner, bug-bounty, and disclosure findings, and retest patches, with Guardian review limiting reachable endpoints. Defense Factory, a ChatGPT preview plugin, traces findings from Daybreak and tools including Wiz, Snyk, and Endor Labs, then drafts patches and GitHub pull requests that engineers must approve. On November 18, Tenable said One customers can import selected code findings locally through OpenAI’s open-source connector.
- Red Team early access validates scanner, bounty, and disclosure findings, then retests patches.
- Guardian review and isolated environments limit which endpoints agents can reach.
- Defense Factory is a ChatGPT preview plugin and does not auto-deploy patches.
- It can use Wiz, Snyk, Endor Labs, and connected GitHub findings.
- The Decoder says both products run on GPT-6 Cyber.
- The November 18 connector imports Tenable One code findings with severity and solution text.
Timeline
dated events in the reporting · oldest first- Nov 12, 2026OpenAI announced both Daybreak tools at Cyber at the Frontier · Cyber Security News
- Nov 18, 2026Tenable described open-source import into Defense Factory · Tenable Blog
What each outlet adds
beyond the shared facts- SecurityWeekPreview plugin, GitHub monitoring, and Wiz, Snyk, and Endor Labs.
- Help Net SecurityIsolated scans, optional Deep mode, and local or cloud checkouts.
- The RegisterBrockman tied security to the frontier model customers use.
- Tenable BlogOpen-source CLI imports Tenable One findings for local review.
- GBHackersEarly-access registration does not guarantee Red Team admission.
Coverage
oldest first · one row per article · UTCOpenAI launches Daybreak Red Team and Defense Factory tools
Early-access agents use GPT-6 Cyber to scan code and propose reviewed patches, while nonprofits sued to halt more capable models.
OpenAI launches Daybreak Red Team and Defense Factory tools
Early-access Red Team tests whether web-app findings are exploitable; Defense Factory drafts patches engineers must approve.
OpenAI unveils Daybreak Red Team and Defense Factory
OpenAI is taking early-access requests for Red Team; Defense Factory is a ChatGPT preview plugin and does not auto-deploy patches.
OpenAI Daybreak adds agents for testing and vulnerability fixes
Early-access Red Team confirms flaws and retests patches, while Defense Factory drafts fixes engineers review before merge.
OpenAI launches Daybreak agents to test and prepare vulnerability fixes
Early-access Daybreak Red Team validates attack paths, while Defense Factory drafts patches that engineers must still review.
OpenAI launches Daybreak Red Team and Defense Factory
Red Team is in early access; both products run on OpenAI’s GPT-6 Cyber model.
Tenable One findings import into OpenAI Codex Security
OpenAI's open-source connector attaches Tenable's description and solution to selected code findings locally.