The Dilemma of AI Implementation: Finding the Right Balance between Productivity and Security
As artificial intelligence (AI) continues to revolutionize various industries, companies are facing a daunting decision: whether or not to implement AI solutions in their operations. On one hand, embracing AI can lead to significant productivity gains, improved efficiency, and enhanced customer experiences. However, on the other hand, implementing AI without proper security measures can expose businesses and their clients to unmitigated risks.
This is where a new wave of "security for AI" startups comes in, promising to mitigate these threats and ensure that AI-powered systems are secure and trustworthy. One such company is Mindgard, a British university spinoff that has developed a novel approach to AI security.
The Opaque Nature of Neural Networks
Professor Peter Garraghan, CEO and CTO of Mindgard, explains the unique challenges posed by neural networks: "AI is still software, so all the cyber risks that you probably heard about also apply to AI." However, he adds, "if you look at the opaque nature and intrinsically random behavior of neural networks and systems," this justifies a new approach.
Mindgard’s Dynamic Application Security Testing for AI (DAST-AI) targets vulnerabilities that can only be detected during runtime. This involves continuous and automated red teaming, simulating attacks based on Mindgard’s threat library. For instance, it can test the robustness of image classifiers against adversarial inputs.
A New Approach to AI Security
Garraghan’s background as a professor and researcher focused on AI security has been instrumental in shaping Mindgard’s technology. The field is rapidly evolving, with new threats emerging as Large Language Models (LLMs) continue to advance. Since ChatGPT didn’t exist when Garraghan entered the field, he sensed that NLP and image models could face new threats.
Mindgard’s ongoing ties to Lancaster University can help the company stay ahead of the curve: the company will automatically own the IP to the work of 18 additional doctorate researchers for the next few years. "There’s no company in the world that gets a deal like this," Garraghan says.
A Commercial Product with Research Roots
While Mindgard has ties to research, it is very much a commercial product already. The company offers a SaaS platform, led by co-founder Steve Street as COO and CRO. Enterprises are a natural client for Mindgard, as are traditional red teamers and pen testers. However, the company also works with AI startups that need to show their customers they do AI risk prevention.
Expansion into the US Market
To meet the growing demand for its services, Mindgard is now announcing an $8 million funding round 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 "building the team, product development, R&D, and all the things you might expect from a startup."
Mindgard plans to expand its presence in the US market, where many of its prospective clients are based. The company has recently appointed a VP of marketing, former Next DLP CMO Fergal Glynn, who is based in Boston.
A Small but Mighty Team
Despite its small headcount of 15 employees, Mindgard is poised for growth, with plans to reach 50 employees within the next year. Garraghan emphasizes that the company’s success is not just about numbers: "We’re a team of experts with a passion for AI security, and we’re committed to making a real impact in this field."
The Future of AI Security
As AI continues to transform industries, the need for robust security measures will only grow. Mindgard’s innovative approach to AI security is a beacon of hope for companies seeking to harness the power of AI while minimizing risks.
By combining cutting-edge research with commercial expertise, Mindgard is well-positioned to address the complex challenges posed by AI. As Garraghan notes, "The future of AI is not just about building intelligent systems – it’s also about making them secure and trustworthy."
Related News
- Instabase raises $100M to help companies process unstructured document data
- Apheris rethinks the AI data bottleneck in life science with federated computing
Subscribe to TechCrunch Daily News for the latest updates on AI, startups, and more.
