On this episode of Actually Intelligent, host Max Thon sits down with James Brear, CEO of Mindgard, and Peter Garraghan, the company’s founder and a leading academic in AI and systems security.
Artificial intelligence is quickly becoming part of the core infrastructure of modern enterprises. The problem? Security hasn’t kept pace. Most cybersecurity tools were built for traditional software—not for AI systems. That gap is opening the door to a new class of vulnerabilities many organizations are only beginning to understand.
James and Peter share how Mindgard emerged from years of academic research into neural network vulnerabilities—long before large language models and AI agents went mainstream. That early work became the foundation for a company focused entirely on securing AI.
If you’re thinking seriously about the future of AI security—from adversarial threats to safeguarding intelligent systems at scale—this episode is worth your time.