Darktrace, today reported that its Cyber AI Analyst product has performed millions of threat investigations, mimicking human thought processes to zoom in on and explore potential threats, and report on the severity of an attack. Mike Beck, Global CISO at Darktrace, will be discussing these capabilities and real-world examples during a presentation at Black Hat USA on Wednesday, August 5th at 10am PDT.
Initially released last September following a 3-year research project at Darktrace’s AI labs in Cambridge, UK, Cyber AI Analyst trains by observing world-class security analysts performing investigations – not on examples of previous attacks. The technology questions raw security alerts, seeks out additional context, and asks questions of third-party sources to come to a conclusion. Using natural language processing, the AI Analyst reports relevant findings at the right moment to security operations personnel, in their local language.
With human security analysts spending an average of 3 hours per security investigation, scaling teams to meet the demand for fast and accurate triaging has become unviable. Today, thousands of organizations rely on Cyber AI Analyst to run investigations alongside their teams, delivering a 92% time saving.