Virtue AI

Cybersecurity · Artificial Intelligence

Acquisition

Fortinet acquires AI runtime security startup Virtue AI

Most of the target's researchers are heading to Meta's Superintelligence Labs while Fortinet keeps the technology.

• Fortinet announced on August 17, 2026 that it acquired Virtue AI, an AI runtime protection and validation startup, for undisclosed terms it called immaterial to its business.
• Founded in 2024 by professors from Stanford, UC Berkeley and University of Illinois Urbana-Champaign, Virtue AI built a Guardian Agent platform for testing and guarding AI models and autonomous agents.
• The roughly 30-person startup's technology is folded into Fortinet's FortiAIGate platform, while most of its employees are instead joining Meta's Superintelligence Labs.
• Virtue AI's platform runs automated red-teaming with over 100 proprietary attack algorithms across hundreds of attack vectors, plus real-time guardrails for text, code, audio, video and image traffic.
• Virtue AI had raised $30 million across seed and Series A rounds in 2025 before being acquired.
• The deal is one of more than 240 cybersecurity M&A transactions tracked so far in 2026, reflecting a wave of consolidation around securing AI agents and models.
• Splitting a startup's talent and technology between an AI lab and a security acquirer shows how fiercely both AI developers and security vendors are competing for scarce AI-safety expertise.