Reverse engineering a sophisticated piece of malware can take a top security researcher four to six weeks. It is painstaking work in which researchers pull apart software to understand how it behaves, often retracing their steps as they uncover new information. Gregor Stewart, Chief AI Officer at Sentinel One, says OpenAI Daybreak can now produce human-level reverse-engineering output in roughly eight hours.
By reducing the time needed for deep analysis, OpenAI gives Sentinel One's researchers room to investigate more malware and connect findings across samples. That wider view often reveals relationships that isolated investigations might miss. Stewart describes the opportunity as "a kind of intelligence that we were never able to afford before."
Daybreak gives defenders access to OpenAI models for cybersecurity work. Sentinel One uses these capabilities to reverse engineer malware, and faster analysis gives its researchers more capacity to investigate additional samples. Stewart emphasizes that his team remains responsible for the outcome. "The only thing that can't be automated is accountability," he says. Decisions to change the architecture still warrant human review, while point fixes may be better suited for automation.
Historically, Sentinel One reserved deep investigation for the most important threats it uncovered. "We were only able to reverse very small numbers of pieces of software," Stewart says. With faster analysis, researchers can map a much larger part of the malware landscape. A component found in two unrelated samples could give researchers a lead during a customer investigation, helping the company defend further out and be more proactive.
Stewart says access to OpenAI's cyber capabilities is allowing more people across the company to take responsibility for security work. Developers may also be able to address security issues as they build software.
Source: OpenAI Blog · Summarized by HeadlinesBriefing