TrendAI Claims Top Spot on CyberGym Security Benchmark

The new benchmark result represents a 3.8 percentage point lead over the previous top-scoring tool, which hit 93.2% earlier this month. TrendAI achieves this by orchestrating seven distinct AI models from providers including Anthropic, DeepSeek, Google, and OpenAI. The system focuses on shrinking the interval between vulnerability discovery and remediation, a window that frequently leaves organizations exposed to active exploitation.

Rachel Jin, Chief Platform and Business Officer at TrendAI, stated that the goal is to render the gap between discovery and protection negligible. By leveraging two decades of research from the company’s ZDI division, the engine generates working proofs of concept for identified weaknesses. This allows security teams to prioritize actual business risk over general noise. Instead of waiting for traditional manual patching cycles, the system deploys virtual patches and real-time threat intelligence to neutralize risks immediately. This proactive defense model aims to transform vulnerability management from a reactive hunt for updates into a continuous, automated shield against live exploitation.