SentinelOne Reports 2,200% Detection Increase Using AI Hunting
SentinelOne reports a massive increase in threat detection through its AI-assisted hunting capabilities and a hybrid human-AI security strategy.
SentinelOne presented a cybersecurity strategy at OneCon 2026 in Las Vegas, emphasizing a hybrid approach that integrates artificial intelligence with human expertise. Chief Customer Officer Steve Stone reported that the company's AI-assisted threat hunting capability, which launched in December 2024, improved the detection of malicious activity by 2,200% over one year. The system currently identifies approximately 28 intrusions per day.
Stone noted that while frontier AI models double offensive cyber capabilities every four months, AI-driven attacks are often easier to identify because they move at machine speed and lack stealth. The company utilizes a seven-layer detection framework and highlighted the React2Shell vulnerability as a primary stress test, where AI-assisted hunting provided the first detection of the flaw. SentinelOne is also using AI to identify evolving delivery wrappers for SectopRAT.
Financially, the company reported 21% growth with $1.1 billion in revenue over the last twelve months. Despite this growth, SentinelOne remains unprofitable, reporting a loss of $1.02 per share.
