Company news
Darktrace makes SECURE AI generally available to monitor enterprise AI use
Published 22 September 2026
WHAT HAPPENED
Darktrace released SECURE AI as a generally available product to monitor how enterprises use artificial intelligence tools. The software detects jailbreak attempts, sensitive data exposure, and unauthorized AI use by scanning sessions in ChatGPT Enterprise, Claude, Microsoft Copilot, and Amazon Bedrock in real time.
WHO IS INVOLVED
Darktrace Holdings Ltd., a U.K.-based cybersecurity company led by President and CEO Ed Jennings, built SECURE AI with integrations from Amazon Web Services, Anthropic, Microsoft, and OpenAI. Darktrace is part of OpenAI's Daybreak Defense Network.
WHY IT MATTERS
Telemetry from 8,200 Darktrace deployments shows more than 80% of monitored customer accounts used generative AI services in August, with the average organization managing five different AI providers. Unmanaged AI use has created security gaps: one customer found employees using unauthorized services despite approving only a single assistant, while another discovered nearly 90 unapproved AI agents built in a low-code environment.
Read more
- Type
- Product launch
- Category
- Security Operations
- Source
- siliconangle.com ↗
