Fortinet adds AI oversight & data controls to FortiEndpoint

Fortinet has expanded FortiEndpoint with new functions for managing AI use, data protection and endpoint risk. The update brings several security features into the company's endpoint platform through a single agent, console and licence.

It adds visibility and controls for AI applications and agents used on endpoint devices, along with built-in data loss prevention, endpoint risk scoring and AI-assisted operational tools. The package is intended to give security teams a more consolidated way to monitor devices, user activity and access decisions across distributed environments.

AI oversight

A central element of the expansion is oversight of both sanctioned and unsanctioned AI tools. FortiEndpoint is designed to show organisations which AI applications, agents and web-based tools are in use on endpoint devices, helping teams identify shadow AI and other unauthorised software.

Security teams can then set policies to allow, restrict, monitor or block those tools according to internal security, compliance and data rules. This puts AI governance at the endpoint rather than relying only on network- or cloud-level controls.

Fortinet presented the release as part of a broader push to simplify security management as companies adopt more AI-enabled software in day-to-day work. It also tied the update to a wider industry problem of fragmented tools across endpoint protection, secure access and data security.

Michael Xie outlined that approach in comments released with the announcement.

"Organisations need a simpler and more effective way to manage security as their environments become more complex and AI-enabled. The Fortinet Security Fabric is designed to converge critical security and networking functions across the enterprise, helping customers reduce complexity, improve visibility, and strengthen protection. With FortiEndpoint, we are extending that strategy by consolidating security, secure access, data security, AI visibility, and assisted operations in a unified endpoint platform, delivered through one agent, one console, and one license," said Michael Xie, Founder, President, and CTO, Fortinet.

Data controls

Fortinet has also added native data loss prevention to FortiEndpoint. The feature is intended to inspect sensitive information exchanged with AI applications, agents and web services at the endpoint, aiming to reduce the risk of data leakage or insider misuse.

The system can provide in-product coaching to users in real time when policies are triggered. The goal is to guide acceptable use of AI tools without blocking routine work in every case.

The company highlighted data categories including personally identifiable information, intellectual property and financial information. By placing those controls in the endpoint platform itself, Fortinet aims to avoid the need for a separate data security product in some use cases.

Operations layer

Another part of the release is FortiAI-Assist, now integrated into FortiEndpoint. The tool lets security staff use natural language prompts to investigate events, generate summaries, identify high-risk devices and troubleshoot issues.

The operational layer also offers contextual insights, policy recommendations and risk guidance to help analysts handle investigations and prioritise threats. These functions sit alongside adaptive zero-trust measures that use changing device health, compliance and risk data to influence access decisions.

That means access to AI applications or protected resources can be tightened or adjusted when an endpoint's risk profile changes. This is intended to support more consistent policy enforcement based on real-time context.

Market pressure

The announcement comes as security suppliers try to respond to a rapid increase in AI use inside businesses, often through a mix of approved and unofficial tools. For many cyber teams, that creates both a governance and data security problem, especially when employees paste sensitive information into external AI services.

At the same time, vendors are increasingly packaging endpoint protection, access control and data safeguards into broader platforms rather than selling them as disconnected products. Fortinet's latest move fits that trend, with an emphasis on fewer management layers and tighter links between endpoint telemetry and other security controls.

Industry analyst Chris DePuy said the integrated approach reflected immediate concerns among security leaders.

"Fortinet is addressing what many CISOs need now: visibility into AI usage, control over sanctioned and unsanctioned tools, protection against sensitive data leakage, and real-time coaching to help employees use AI responsibly. Delivering these capabilities through FortiEndpoint gives customers a practical way to manage AI risk with the same agent and licence they already rely on for endpoint security," said DePuy.

The new FortiEndpoint enhancements are expected to become available in the third quarter.