HiddenLayer, a security provider for AI systems, has joined the Databricks Unity AI Gateway ecosystem to extend model security and threat detection capabilities into Databricks-governed AI workflows. The collaboration builds on HiddenLayer's existing Unity Catalog integration. The partnership addresses security challenges as enterprises move from AI experimentation to production deployment of generative AI, agents and Model Context Protocol integrations. HiddenLayer will provide AI-native security intelligence, runtime monitoring and detection-and-response capabilities within Databricks' platform. The integration offers model security scanning, runtime visibility for AI workloads, threat detection for attacks like prompt injection and data leakage, and policy controls. Founded to address AI-specific security threats, HiddenLayer protects AI applications across the entire lifecycle, from supply chain security to runtime protection.

HiddenLayer, an AI security company, has partnered with Cohere to enable enterprises to deploy agentic AI at scale with built-in protection against AI-native threats. The collaboration combines Cohere's North platform with HiddenLayer's AI Security Platform to address risks including prompt injection, data leakage and malicious tool use. The partnership allows organisations to deploy AI agents that connect to enterprise systems whilst defending against attacks that traditional security tools cannot address. HiddenLayer tested the solution internally before deployment, demonstrating practical security for production environments. Key capabilities include detecting prompt injection and model attacks at runtime, reducing data leakage risks, and providing audit-ready visibility for compliance. The companies will host a webinar on 30th June to discuss secure enterprise AI deployment.