NeuralTrust Debuts Runtime Security Mesh for AI Agent Ecosystems

Until now, securing AI agents required either siloed point solutions for specific tools or the constant rebuilding of integrations as teams adopted new technologies. NeuralTrust aims to solve this fragmentation by deploying an Agent Gateway that sits at the center of agentic traffic. This approach moves beyond checking system prompts or permitted tool lists, instead tracking an agent's entire operational loop.

The platform validates every stage of an agent's activity: incoming prompts, tool calls, external responses, and the agent's internal reasoning process. By monitoring these cycles, the software identifies potential malicious database queries, indirect prompt injections, and unauthorized data access before they trigger action. According to Joan Vendrell, co-founder and CEO, the company developed this runtime security to address a market that refuses to consolidate around a single AI provider.

NeuralTrust’s approach has garnered industry attention, including recognition as a Representative Vendor in multiple Gartner Market Guides and a leadership position in the 2025 KuppingerCole Leadership Compass for Generative AI Defense. The company is currently demonstrating its TrustGate and TrustGuard technologies at Black Hat USA 2026, targeting enterprise clients across the banking, airline, and energy sectors.