Block uses OpenAI Daybreak to boost security triage throughput by 50%

The implementation has yielded measurable efficiency gains. Block reports a 50% increase in triage throughput and a 50% rise in intelligence gathering and investigations. The system now completes approximately 1,100 analyst-hours of investigation per quarter, while returning roughly 600 analyst-hours previously spent on false-positive triage. In recent red team exercises, the time from identifying a high-level vulnerability to generating an automated proposed patch was reduced to 18 minutes.

James Nettesheim, Block’s Chief Information Security Officer, emphasizes that the AI acts as a scalable teammate rather than a simple automation tool. The system routes tasks to different models based on the job, with humans validating results and providing organizational context. Block plans to gradually increase the risk level of AI-assisted tasks as the technology’s judgment is proven in practice, aiming to reduce the number of vulnerabilities introduced in the first place.