Integrating Forcepoint AI Security with Forcepoint DLP On-Prem

Important: The features described in this chapter are available only from Forcepoint DLP 10.4.5. If you are running an earlier 10.4 release, the related pages and options are not present in Forcepoint Security Manager. These features also require a Forcepoint AI Security license, a provisioned Data Protection Service, and a DLP license that provides access to the DLP policies in the Data module of Forcepoint Security Manager (DLP Endpoint, DLP Suite, or DLP Network).

Forcepoint DLP extends the data security policies you already maintain to the AI channels the organization uses. The same classifiers, policies, and action plans that protect email, web, and cloud application traffic can be applied to the prompts the users send to AI applications and to the requests the AI agents make against the business systems.

Forcepoint DLP supports the following two pillars of Forcepoint AI Security:

AI Security (AI Detection & Response) over DLP Cloud API channel: Governs sanctioned AI. It detects when sensitive data leaves the organization through AI applications and can monitor and mitigate users prompts and AI responses. This solution is monitored by selecting AI applications to be detected over the DLP Cloud API channel in DLP policies. See Getting started with AI Security (AI Detection & Response) over DLP Cloud API channel.
MCP Gateway (AI Agent Gateway): AI Agents can call business applications like Atlassian exactly as an employee would, bypassing DLP inspection and leaving no audit trail. Forcepoint MCP Gateway connectors sit between every agent and the applications it calls. Every request is inspected before it executes, and sensitive data is about to leave the organization. Now DLP administrator can configure DLP policies using the new MCP Gateway channel. See Getting started with the MCP Gateway (AI Agent Gateway) channel.