Last Updated: 12 August 2026
Mindgard, a startup that helps organizations find and fix security weaknesses in their AI systems, has raised $30 million in a Series A round. Founded by Dr. Peter Garraghan, a researcher at Lancaster University in the UK, the company builds software that simulates real-world attacks on AI models, agents, and applications. The approach is known as "red teaming," essentially stress testing AI systems to find vulnerabilities before real attackers do.
The funding comes at a time when companies across industries are racing to put AI into production, often without a clear way to secure it. AI systems can be tricked into leaking sensitive data, bypassing safety rules, or executing malicious code through techniques like prompt injection and adversarial attacks. Traditional security tools were not built for these risks. The broader AI security platforms market was valued at $13.2 billion in 2025 and is projected to reach $56.5 billion by 2033, a sign of how quickly demand is growing.
Album VC led the round, with participation from Karma Ventures and returning investors .406 Ventures, Atlantic Bridge, IQ Capital, and Lakestar.
AI security differs from conventional cybersecurity in a fundamental way. Traditional software is deterministic: give it the same input and it produces the same output. AI models are probabilistic. They behave in ways that are often opaque even to the people who built them. Mindgard calls this the "psycho-technical attack surface," meaning attackers can manipulate AI systems by exploiting learned behaviors rather than conventional software bugs.
The company has publicly disclosed more than 150 high-impact security and safety vulnerabilities across widely used AI products. Those findings include a zero-day code execution flaw in Cursor IDE, a popular AI coding tool, a trusted workspace flaw in Google Antigravity, and failures in ChatGPT's image generation safety guardrails. Each vulnerability feeds back into the company's knowledge base, strengthening its testing and defense over time.
"Organizations are moving AI into critical operations without security infrastructure designed for how these systems operate in practice. Mindgard has translated deep research and elite offensive-security expertise into a continuously evolving capability that enables enterprises to stay ahead of emerging AI threats. We believe this team can define the AI security category."
Ty Boswell, Partner at Album VC
Mindgard will put the new capital toward product development, engineering, sales, and marketing. The company says significant customer demand is driving the expansion. Its priority is the United States, where much of its Fortune 2000 customer base is concentrated.
The round follows what Mindgard describes as accelerated deployment across a large number of Fortune 2000 companies and AI innovators in financial services, pharmaceuticals, gaming, digital services, semiconductors, and healthcare. As more organizations bring AI into production environments, the need for specialized security testing is growing alongside it.
"AI is creating an entirely new attack surface and organizations need a fundamentally different approach to securing it. We don't just automate attacks. We operationalize expertise, turning the knowledge of leading AI security researchers and offensive security practitioners into the capabilities every enterprise needs to secure their AI. This investment will expand our global reach and help make attacker-driven AI security a core part of how organizations implement and manage AI."
James Brear, CEO of Mindgard
Mindgard was founded in 2022 by Dr. Peter Garraghan, along with co-founders Steve Street and Dr. Neeraj Suri. It spun out of Lancaster University, where Dr. Garraghan led what the company describes as the world's largest AI security lab. Today it is headquartered in Boston and London.
The platform does three things. It discovers "shadow AI," meaning AI tools and models employees may be using without the security team's knowledge. It performs AI red teaming, simulating attacks against AI systems to uncover weaknesses. And it provides runtime protection, monitoring AI systems in real time and responding to threats as they happen. Security teams can use it without needing specialized AI expertise because it integrates with existing enterprise tools and workflows.
CEO James Brear leads the company, with Dr. Garraghan serving as Chief Science Officer. Aaron Portnoy joined as Chief Product Officer and Rich Smith as VP of Research, both in 2025, to strengthen offensive security capabilities.
Album VC, founded in 2014 and based in Lehi, Utah, led the $30 million round. The firm focuses on early-stage technology companies. Karma Ventures, a European venture firm, participated as a new investor.
All four existing backers returned. .406 Ventures led the company's previous $8 million round in early 2025. IQ Capital and Lakestar were the original seed investors, backing Mindgard's £3 million round in 2023. Atlantic Bridge also participated in earlier funding. With the Series A included, Mindgard's total disclosed funding comes to approximately $41 million.
"AI security is an emerging category and Mindgard stands out as one of the names to watch. Its technology is built on world-class offensive security research, combining automated reconnaissance with smart penetration testing. Crucially, it’s already proven in production - some of the most demanding F2000 enterprises rely on it to safeguard their AI systems. That combination of a defensible research moat and proven enterprise pull is rare. We’re proud to support James, Peter, and the team in defining this category."
Kristjan Laanemaa, Founding Partner at Karma Ventures
