Home
Latest reporting
News
Cisco says its network intelligence and OpenAI models are being combined to improve enterprise security and detect threats more adaptively.
Cisco and OpenAI are describing a security approach that combines network intelligence with reasoning models to help enterprises detect threats more adaptively, according to OpenAI’s Cisco cyber story published on 3 November 2026. The companies say the aim is to correlate signals that may look benign on their own but reveal malicious activity when assessed together.
Evidence and context
OpenAI says the Cisco deployment pairs Cisco’s network intelligence with OpenAI models to strengthen enterprise security and operate at what Cisco President and Chief Product Officer Jeetu Patel describes as “machine speed and scale.” The same material says the system uses security telemetry alongside observability data so defenders can distinguish between hostile interference and autonomous agents that are pursuing legitimate goals in unintended ways.
OpenAI also says the deployment is supported by zero data retention, customer-data boundaries and isolation. Those arrangements are described in the source as part of this deployment and should not be assumed to apply to all OpenAI services.
Related OpenAI security work
The Cisco story follows OpenAI’s 2 November 2026 post, “Putting frontier cyber models in more trusted hands,” which describes the broader Daybreak Cyber Partner Program for governed access to defensive workflows and specialized red-team and penetration-testing work. A separate OpenAI customer story dated 27 May 2026 says Cisco has also used Codex in engineering workflows, including work on Cisco AI Defense, to accelerate compile-test-fix cycles and defect remediation.
The available material is limited to company descriptions and linked customer stories. It does not provide independent performance testing, quantitative security outcomes or a broader benchmark comparing Cisco’s approach with other enterprise security systems.
Be the first to comment