Forcepoint adds Claude oversight for enterprise AI

Forcepoint has integrated its security platform with the Claude Compliance API, bringing Claude Enterprise into its existing oversight tools for workplace AI systems.

The integration feeds Claude Enterprise conversations, file uploads, generated responses and user activity into a single dashboard for security and compliance teams. Customers can monitor the use of sensitive information and apply controls across Claude alongside Microsoft 365 Copilot and ChatGPT Enterprise.

The announcement reflects growing demand from large companies for tighter oversight of generative AI tools used in day-to-day work. As these systems are applied to email, source code, financial data and customer records, businesses are under pressure to track what information goes into the tools, what comes back out, and whether any regulated material is exposed.

When the service is first connected, historical Claude activity is imported, giving security teams an existing record rather than forcing them to start from scratch. The console also covers approved AI tools, unauthorised or "shadow" AI use, and real-time inspection of prompts and responses for supported applications.

Compliance focus

A central part of the rollout is compliance reporting. The system is designed to help customers gather evidence for AI governance obligations, including requirements linked to the EU AI Act, the NIST AI Risk Management Framework and US Securities and Exchange Commission disclosure expectations around AI.

The platform also applies the same policy framework across cloud services, websites, networks, email and endpoints. The aim is to let organisations classify data once and enforce the same rules when that data is used inside AI systems.

Vendors in security and governance software have been moving quickly to position themselves around enterprise AI adoption, particularly as tools from Anthropic, Microsoft and OpenAI gain wider use in regulated sectors. For customers, the challenge is less whether employees are using AI than whether existing controls can see that activity in enough detail to manage legal, operational and reputational risk.

For Forcepoint, the Claude integration extends a broader push to present AI monitoring as an extension of data security rather than a separate category. Customers can identify, tag and protect intellectual property and regulated information before it reaches an AI tool, while also detecting data exposure across multiple channels from a single platform.

Ryan Windham, Chief Executive Officer of Forcepoint, set out that position in the company's announcement. "Forcepoint leads the AI security market from the position we've always owned, which is the data. AI changed how information behaves. The enterprises that win will be the ones that start with the data and build up from there - fast enough to keep pace with the AI," Windham said.

He added: "Now we're bringing Claude Enterprise under the same AI governance and data security trusted everywhere else. That gives our customers the ability to say yes to AI in their business, with the controls, the attribution and the proof to back it up. We're stopping what others can't."

Wider rollout

The integration is available to Forcepoint customers that use Claude Enterprise and Claude Platform. By linking Claude activity with records from other AI services, Forcepoint aims to give security teams one place to review usage patterns, investigate incidents and apply policies across systems that might otherwise be monitored separately.

That matters because enterprise AI deployments are becoming harder to separate into neat categories. Companies may use one sanctioned assistant for office productivity, another for software development, and a mix of public tools accessed without formal approval. Security teams then face a fragmented view of how staff are handling corporate data.

Forcepoint's approach is to treat AI interactions as another route through which information travels, alongside email, web traffic and cloud applications. In practice, that means logging what users submit to Claude, what files they upload, what responses are generated and how those interactions relate to internal data policies.

The company operates from Austin, Texas, and says it serves customers in more than 150 countries. Its latest integration underlines how AI oversight is shifting from a niche concern to a standard requirement for businesses trying to use generative tools without losing visibility over sensitive data.

Historical Claude activity also loads automatically on first connection, so security teams begin with context rather than a blank slate.