Trent AI, a startup building security infrastructure for autonomous AI systems, has raised $13 million in seed funding and officially emerged from stealth. The round was led by LocalGlobe and Cambridge Innovation Capital, with notable angel investors including Joaquin Quiñonero Candela of OpenAI, Avinash Bhat of AWS, Ippokratis Pandis from Databricks, and former Spotify VP of Engineering and Head of AI/ML Tony Jebara.
The seed capital will go toward expanding Trent AI's engineering team, accelerating the development of its security agents, and growing its design partner and early customer base.
The numbers tell a story the industry is only starting to confront. According to Deloitte's 2026 State of AI report, nearly three in four companies plan to deploy agentic AI within two years. Yet only one in five say they have a mature governance model in place for those systems.
That gap creates real risk. In complex environments where multiple AI agents are interconnected, a single security weakness can expose an entire infrastructure. Traditional security tools were designed for static applications with periodic scanning cycles. They were not built for systems that change continuously and make decisions autonomously.
"Organizations are deploying AI agents and autonomous workflows faster than their security can adapt, and most development teams using these agents and workflows have no security framework designed for their systems," said Eno Thereska, Co-founder and CEO of Trent AI. "This is not an easy problem to solve. Trent AI is tackling these difficult and important problems, while building the necessary security foundations and frameworks for agentic systems now and through the next decade."
Trent AI's platform is built for engineering and security teams that need to ship agentic systems quickly without sacrificing safety. The solution operates across the full development lifecycle using four types of specialized agents.
Scanning agents monitor code, infrastructure, dependencies, and runtime behavior in real time, building a continuous picture of where risk lives in a given system. Judgment agents then classify what matters, separate signal from noise, and prioritize threats based on actual business impact rather than static rule sets. That judgment improves over time as the system learns.
Remediation agents handle the response side: patching vulnerabilities, opening pull requests, adjusting configurations, and verifying that fixes hold. Finally, security posture agents track trends over time, benchmark against industry standards, and surface systemic weaknesses. As each feedback cycle completes, the system becomes more accurate at forecasting risk. Early design partners including Canopy, Commscentre, ML@Cam, Qbeast, and Weblogic have already reported immediate improvements in security visibility, faster vulnerability identification, and clearer remediation scope.
Trent AI was founded in 2025 by three co-founders with complementary backgrounds that are genuinely well-suited to the problem. Eno Thereska is a former Distinguished Engineer at Alcion, acquired by Veeam, with additional experience at AWS and Confluent. Neil Lawrence is DeepMind Professor of Machine Learning at the University of Cambridge and previously served as Director of Machine Learning at Amazon. Zhenwen Dai was a Machine Learning Scientist at AWS before becoming Senior Research Manager at Spotify.
The company is also an active participant in the broader security community. It holds Partner Startup membership with OWASP, the Open Worldwide Application Security Project, is a Startup Partner with Carnegie Mellon University's CyLab Venture Network, and is contributing a Security Agent to open-source platforms including OpenClaw.
"AI models have led to an exponential growth in code being generated by companies big and small. That code brings along an exponential growth in security risks, vulnerabilities and threats... and human security teams just can't keep up," said Tony Jebara, former Spotify VP Engineering and Head of AI/ML. "We desperately need specialized AI models that can analyze this flood of code, produce security assessments and provide mitigations. Trent AI is providing just that."
LocalGlobe and Cambridge Innovation Capital anchored the round, joined by a group of senior technology executives with direct experience building infrastructure at scale.
"Agent adoption is outpacing enterprise security readiness. As autonomous workflows make decisions across critical systems, a new layer of infrastructure is needed to govern, observe, and enforce safe behavior," said Ian Lane, Partner at Cambridge Innovation Capital. "We believe Trent AI is well placed to define this category."
Saul Klein, Co-founder and Executive Chairman of Phoenix Court, the home of LocalGlobe, put the timing plainly. "Now is the right time to build the long-term foundations of security for agentic systems. Trent AI is uniquely positioned to do this, combining deep academic expertise with real-world experience building large-scale systems and working closely with design partners deploying agents today."
Avinash Bhat, Director at AWS and one of the angel investors, added that Trent AI is building the foundations teams will need to operate autonomous systems safely at scale.
