NeuralTrust

NeuralTrust, founded in 2022 in Barcelona, builds security infrastructure specifically for enterprises deploying large language models and AI agents. Its core product is a Generative Application Firewall (GAF) that sits in front of LLM traffic, inspecting and sanitizing prompts and responses to block threats like prompt injection, jailbreak attempts, and sensitive-data exfiltration before they reach or leave the model. Around that firewall, the company layers security agents that enforce permission boundaries on what an AI agent is allowed to do, plus tracing and analytics tooling so security teams can monitor AI behavior and produce compliance evidence rather than treating LLM traffic as a black box.

In June 2026, NeuralTrust raised a $20 million seed round led by German investor Alstin Capital, with participation from VentureFriends, Seaya, Kibo Ventures, Banc Sabadell, Plug and Play’s EA Ventures fund, and IESE Business School’s Finaves fund, plus non-dilutive backing from the European Innovation Council and Spain’s state research agency (AEI). Multiple outlets, including the company’s own announcement and independent coverage from EU-Startups and SiliconANGLE-adjacent press, describe this as the largest cybersecurity seed round raised by an EU company to date — a notable claim given it is a seed round rather than a later-stage raise, and one that should be read as a European-market superlative rather than a global one.

The operating thesis is straightforward: as enterprises move from experimenting with chatbots to deploying autonomous AI agents with real permissions and API access, the attack surface shifts from ‘can a user jailbreak the chatbot’ to ‘can an agent be manipulated into taking an unauthorized action with real consequences.’ NeuralTrust’s product bet is that this requires purpose-built inline controls rather than retrofitting a traditional web application firewall or API gateway to understand LLM-specific threats.