Mindgard Raises $30 Million to Test AI Systems for Security Flaws
The London-and-Boston startup has already found more than 150 vulnerabilities in popular AI products, including a previously unknown flaw in a widely used code editor. Fresh capital will expand its engineering and sales teams.
Key points
- Mindgard closed a $30 million Series A funding round on 22 July 2025, bringing its total funding to roughly $42 million.
- Album VC led the round, with Karma Ventures and four existing investors joining.
- The company says its platform has uncovered more than 150 security flaws across popular AI products since its founding in 2022.
- Discovered vulnerabilities include a zero-day (a flaw the software maker did not previously know about) in Cursor IDE, a popular tool developers use to write code with AI assistance.
- Mindgard plans to use the money to grow its teams and expand into financial services, healthcare, pharmaceutical, and semiconductor sectors.
AI security startup Mindgard has raised $30 million in a Series A round, the company announced today. Founded in 2022 out of Lancaster University, the firm now operates from London and Boston and builds tools that attack AI systems deliberately, to find weaknesses before real criminals do.
What does Mindgard actually do?
Mindgard runs what the security industry calls red-teaming: it plays the role of an attacker, probing a company's AI products for flaws so those flaws can be fixed. The platform automates that process, meaning it can test an AI system continuously rather than in a one-off annual audit.
The company says its software maps every possible way an attacker could manipulate an AI model or application, then keeps watching that surface for new weak spots. When it finds a problem, it also provides guidance on how to block it.
Since launch, the platform has flagged more than 150 vulnerabilities in real AI products. Among them: a zero-day code-execution flaw in Cursor IDE (a code-writing tool that uses AI to help programmers), and security defects in Google Antigravity and ChatGPT.
Who put up the money?
Album VC led the Series A. Karma Ventures joined as a new backer. Four previous investors, .406 Ventures, Atlantic Bridge, IQ Capital, and Lakestar, also participated. Combined with earlier funding, Mindgard has now raised close to $42 million in total.
The company says it will use the capital to hire across product, engineering, sales, and marketing. Target sectors include financial services, gaming, healthcare, pharmaceuticals, and semiconductors.
"AI is creating an entirely new attack surface and organisations need a fundamentally different approach to securing it," Mindgard CEO James Brear said in the announcement, first covered by SecurityWeek.
Why does this matter to ordinary people?
Most people never interact with Mindgard directly. The company sells to businesses, not consumers. But those businesses use AI systems to handle mortgage applications, medical records, customer service chats, and more.
An AI system that has not been tested for security flaws could be manipulated by a criminal to produce wrong outputs, leak private data, or be tricked into bypassing safety rules. Mindgard's pitch is that it catches those problems before they become headlines.
For now, the practical implication for individuals is simple: if a company you use is deploying AI features in a sensitive area, it is worth asking whether that system has been independently tested for security weaknesses.
