British university spinoff Mindgard protects companies from artificial intelligence threats

AI poses a dilemma for companies: Don’t implement it yet, and you could miss out on productivity and other potential benefits; however, getting it wrong can expose your business and your customers to unmitigated risks. This is where a new wave of “security for AI” startups is emerging, with the goal of ignoring these threats like jailbreak and operative injection.

As an Israeli startup Or and US-based competitors Hidden layer and Protect the AIBritish university spinoff Mindguard is one of them. “AI is still software, so all the cyber risks you’re probably hearing about apply to AI,” said its CEO and CTO, Professor Peter Garraghan (pictured right above). But, he added, “if you look at the opaque nature of neural networks and systems and their random behavior,” that justifies the new approach.

In Mindgard’s case, the approach in question is Dynamic Application Security Testing for AI (DAST-AI), which only targets vulnerabilities that can be detected at runtime. It involves continuous and automated red teaming to simulate attacks based on Mindgard’s threat library. For example, it can test the robustness of image classification against competing entries.

On this front and beyond, Mindgard’s technology owes much to Garraghan’s background as a professor and researcher focused on AI security. The field is evolving quickly — ChatGPT didn’t exist when he got into it, but felt that NLP and image models could face new threats, he told TechCrunch.

Since then, he has been looking forward to what the future holds has become a reality within a rapidly growing sector, but LLMs continue to change as do threats. Garraghan believes its ongoing relationship with Lancaster University could help the company stay afloat: Mindgard will automatically own the IP for the work of 18 additional PhD researchers over the next few years. “There is no company in the world that has such a deal.”

Although connected to research, Mindgard is now a commercial product and, more specifically, a SaaS platform, with co-founder Steve Street leading the charge as COO and CRO. (An early co-founder, Neeraj Suri, who was involved on the research side, is no longer with the company.)

Enterprises are a natural customer for Mindgard, such as traditional red teams and pen testers, Garraghan said, but the company also works with AI startups that need to show their customers how to prevent AI risk.

Since many of these potential customers are in the United States, the company added an American flavor to its cover table. after growing a £3 million seed round In 2023, Mindgard is now announcing a new round of $8 million led by Boston-based .406 Ventures with participation from Atlantic Bridge, WillowTree Investments and existing investors IQ Capital and Lakestar.

The funding will help with “team building, product development, R&D and everything else you’d expect from a startup,” while expanding across the US. Boston. However, the company also plans to keep R&D and engineering in London.

With a staff of 15, Mindgard’s team is relatively small and will remain so, with plans to grow to 20-25 by the end of next year. That’s because AI security is “not yet in its heyday.” But when AI becomes ubiquitous and security threats follow suit, Mindgard will be ready. “We built this company to do positive things for the world, and the positive side here is to make sure people can use AI in a safe and secure way,” says Garraghan.