Fortinet adds AI prompt checks to the firewalls its customers have
The new service checks prompts to AI models when they go, and masks sensitive data in them.
Fortinet launched the FortiGuard Inline AI Security Service. It extends the FortiGate firewalls of Fortinet, a network security company, into AI security. The service comes from the cloud of Fortinet. Customers use it on firewalls they have, and they do not use a different AI security stack. Fortinet also added features to FortiAIGate, a gateway between software and AI models.
How it works
The service checks prompts that go to large language models, agents and Model Context Protocol tools. This protocol lets agents use other tools. Fortinet says that used tools look at IP addresses and signatures. A prompt with an instruction for an agent goes through.
The numbers
Fortinet says the data loss prevention engine finds 35 types of private data. The gateway has more than 50 compliance policies in it. Fortinet gave no test results from an external group.
What is not known
Fortinet gives no price. In August, Fortinet bought Virtue AI, a security startup. Analyst Will Townsend of LoneStar Advisory and Research says that the fast use of this technology improved the guardrails.
Insight: AI security is moving into firewalls that customers have
Fortinet sells the new service on firewalls that customers have. It bought Virtue AI in August, and that technology is in the guardrails of the gateway. This reads as a plan to keep AI security in the firewall budget. A different tool can win that spend if it does not. Look for the same step from other firewall vendors.
Sources
Posted