In the AI: Trust but Verify podcast, our host, Alec Crawford (@alec06830), Founder and CEO of Verapath (https://www.verapath.com), interviews guests about how they are using AI in business, where you can trust AI, and where you need to put up guard rails.

In this episode of AI: Trust but Verify, Alec Crawford sits down with Omid Pakseresht, CEO of Goodfolio (renamed Sepanta), to discuss what it takes to build trustworthy AI systems in a world where organizations are rapidly adopting AI. The conversation explores why accuracy alone is not enough, the importance of governance and guardrails, and how AI often exposes weaknesses in existing business processes, data quality, and organizational controls.

Alec and Omid also discuss AI's growing role in cybersecurity, the challenge of managing model costs, the future of AGI, and whether AI's greatest value will come from efficiency gains or entirely new opportunities for growth. While both acknowledge the risks associated with increasingly powerful AI systems, they remain optimistic that organizations can use AI responsibly if they focus on trust, verification, and human judgment.

Top 3 Takeaways

1. Don't trust AI with high-impact decisions without guardrails. In areas like finance and healthcare, AI should support human decision-making, not replace it.
2. Accuracy is less important than system design. Governance, controls, testing, and verification processes matter more than simply choosing the best model.
3. AI's biggest opportunity may be growth, not cost reduction. The most transformative use cases will likely come from creating new capabilities and business models rather than simply making existing processes more efficient.

Mentioned

Garry Kasparov

Jimmy Stewart

Elon Musk

Books

Risk Management in the AI Era (forthcoming) by Alec Crawford