Fastly, Inc., based in San Francisco, has unveiled three new AI security and control capabilities as part of its Fastly for AI portfolio. Launched on Monday, September 21, 2026, the tools—AI Runtime Control, AI Firewall, and enhanced API Security—aim to provide enterprises with granular oversight over AI systems deployed on its edge cloud platform. These additions follow a surge in machine-generated traffic, which exceeded 50% of total requests across Fastly’s network in July and August 2026, with AI traffic growing 6.5 times faster than human traffic from January through May 2026. According to McKinsey research, 93% of organizations report exceeding their AI budgets, highlighting the need for tighter controls to mitigate risks and optimize costs.
The AI Runtime Control feature consolidates model interactions through a unified endpoint, offering visibility into token usage, rate limiting, budget management, and failover mechanisms. The AI Firewall evaluates incoming requests to detect and block potential prompt injection attacks, while the enhanced API Security module enforces API contracts and monitors or blocks non-compliant requests. These tools integrate seamlessly with Fastly’s existing platform features, including Bot Management, DDoS Protection, and API Discovery, reinforcing the company’s edge network capabilities. As of June 30, 2026, Fastly’s global edge network handled over 622 terabits per second (Tbps) of traffic, with daily request volumes exceeding five trillion as of March 31, 2026.
Kelly Shortridge, Fastly’s Chief Product Officer, emphasized the urgency for enterprises to implement runtime controls for AI adoption without disrupting operational velocity or compromising existing security resilience. The new suite reflects the accelerating shift toward AI-driven applications, underscoring the need for robust, real-time security measures to safeguard digital infrastructure amid rapid technological evolution.
