OpenAI and Cogent say they are building "cyber defense factories" for enterprises: systems meant to fix security vulnerabilities automatically.
According to OpenAI's post, titled "Cogent and OpenAI build cyber defense factories for enterprises", Cogent combines OpenAI models with enterprise context to help automatically remediate vulnerabilities with agents. Remediate means fix. Agents are AI programs that carry out multi-step tasks on their own rather than answering a single prompt. The summary of the post available to us does not name customers, give counts of flaws fixed, or say whether a person reviews each change.
The problem is a familiar one. Scanners find flaws faster than security teams can patch them, so the backlog grows and humans triage it by hand. Fixing is the slow part, which is where this pitch aims.
The "enterprise context" phrase is the piece to watch. A patch that is correct in the abstract can still break a particular company's systems, so an agent that knows what runs where has a better chance of producing a fix that holds. That is an inference from the wording, not something the summary spells out.
The idea is not new. GitHub's Copilot Autofix suggests fixes for code-scanning alerts, and Google has described its own CodeMender agent for the same job. The "factory" framing suggests volume, with fixes produced as a pipeline rather than one at a time.
The open question is oversight. Automated fixes are easy to accept in a test environment and much harder to trust in production, and how much sign-off these agents require will say more than the branding does.