Darktrace Launches AI Security Monitor for Enterprises

Darktrace has made its SECURE AI product generally available, using behavioral detection to monitor how employees and contractors use AI tools across the enterprise. Telemetry from 8,200 deployments shows over 80% of customers used generative AI in August, with the average organization relying on five different AI providers. The platform integrates with AWS, Microsoft, Anthropic and OpenAI, scanning sessions in ChatGPT Enterprise, Claude, Copilot and Bedrock for jailbreaks, data leaks and prompt injection in real time. It also detects shadow AI, maps agents to human users, and checks company AI policies against compliance frameworks.