Opsin Integrates with Claude's Compliance API to Extend AI Governance to Anthropic-Hosted Deployments

Opsin announced an integration with Claude's Compliance API, extending Opsin's governance capabilities to organizations running Claude, powered by Anthropic. The integration is listed on the Anthropic website and is available now for enterprises across industries.

As enterprises deploy Claude across business functions, security and compliance teams face a growing gap between AI adoption and governance coverage. Claude is widely used across regulated industries, and organizations have lacked a way to bring that activity into the same governance framework as their other enterprise AI tools.

The integration connects Opsin's platform to Claude's Compliance API, pulling in activity from Anthropic-hosted deployments - including Claude Enterprise - so security teams can monitor how Claude is being used and flag risk as it happens.

For Claude Enterprise deployments, Opsin can also bring in conversation-level context, giving teams visibility into how agents and users are interacting with enterprise data and surfacing a risk signal when something looks off. The integration applies to Anthropic-hosted deployments and does not extend to Claude on Amazon Bedrock or Google Cloud's Vertex AI.

For regulated enterprises, this means Claude activity is no longer a blind spot. Opsin customers can now apply the same access, behavior, and data governance policies to Claude that they apply across the rest of their sanctioned AI environment.

Opsin has reduced sensitive data exposure by an average of 81% across its customer base, and the Claude integration extends that coverage to one of the most widely deployed enterprise AI products.

"Regulated enterprises aren't choosing between Claude and governance. They need both, and until now they've had to manage them separately," said James Pham, CEO, Opsin. "This integration closes that gap. Security teams can see what's happening across Claude deployments, connect it to identity and data context, and respond before risk becomes an incident."

Source: Opsin media announcement