Fortinet recently held its virtual AI cybersecurity summit, offering participants practical advice on securing AI, using it to strengthen defensive posture, and prioritizing network and security operations. I found two sessions particularly interesting, one related to mitigating shadow AI and a second that focuses on the need to reimagine endpoint AI security.
Mitigating Shadow AI
At RSA Conference 2026, the need for deeper agentic observability dominated many of my conversations, so it was no surprise that shadow AI would be a subject of discussion at the Fortinet virtual event. I agree with the company’s assessment that organizations are dealing with a looming crisis. As AI experimentation continues, there are large groups of knowledge workers using AI tools and models and deploying agents without proper oversight – and the potential damage is catastrophic, especially in highly regulated industries including healthcare and financial services. Furthermore, the use of intellectual property and proprietary code bases increases the likelihood of data leakage with AI coding assistants.
To address the concerns arising from shadow AI, Fortinet is positioning FortiOS 8.0 as its tip of the spear in improving organizational AI visibility. It does so by facilitating native visibility into AI application usage through application control and deep inspection. The granularity is particularly compelling, allowing security teams to identify specific AI services, by whom, and in what context. Most importantly, it can be done without the need for additional tools or configuration settings for customers who have adopted the converged networking and security fabric. FortiGuard Labs also enriches insights through a continual classification of emerging service and usage patterns, and its complement to FortiOS makes the pair a powerful combination in mitigating and eventually eliminating shadow AI.
Reimagining Endpoint AI Security
The adoption of AI is moving at light speed, and in addition to the need to eliminate the negative consequences tied to shadow AI, Fortinet recognized during its event that there is an equal requirement to reimagine endpoint AI security. With the rapid rise of experimentation tied to OpenClaw, now agents can be created and reside on any endpoint within a network, and that poses a significant risk to organizations of all sizes.
Fortinet has wisely anchored its endpoint security services to its FortiGate firewall platform, and this positions it as an “easy button” to address what is required for endpoint AI security. Features can be easily toggled on, including FortiDLP, to safeguard against data leakage, and FortiSASE ensures that AI usage is governed, regardless of where it occurs within the expanse of an infrastructure estate. When combined with Fortinet’s next-generation firewall capabilities that include application control, web filtering, DLP, and SSL inspection, coupled with FortiOS’s fabric that also includes MCP support for agentic frameworks, organizations can take advantage of a deep set of capabilities that address the need for stronger endpoint AI security control.
Final Thoughts
Fortinet is often considered a value player within networking and security infrastructure and services. It is not a knock on the company, and it could certainly charge more for its broad and deep portfolio that boasts custom, purpose-built silicon and a robust fabric.
What is noteworthy is that Fortinet has architected an optimal tech stack over the past two and a half decades that squarely addresses security operational needs tied to modern AI applications and usage. Did the company’s founders have a crystal ball? Maybe not, but a relentless pursuit of delivering customer value has clearly paid dividends and positioned it well to capitalize on the need to secure modern AI at scale.
