Building Trust with AI Agents - Henri Terho

📘 Free e-book: The 7 success factors of software testing. 25 years of project experience in one 33-page workbook, now also in English 👉 Get it for free

"AI doesn't think, it doesn't analyze, it predicts." - Henri Terho

In this episode, I talk with Henri Terho, senior consultant and AI enthusiast, about why building trust in AI systems requires the same rigor we've always applied to software—just now at a whole new level. Henri explains how AI agents multiply both our successes and our mistakes, why prompting is harder than it looks, and why testers are uniquely positioned to thrive in this shift. We dig into the oracle problem, the communication trap, and why your test suite might soon matter more than your codebase.

Henri Terho is a Senior AI Consultant at Eficode with broad experience spanning regulated industries—automotive, banking, aerospace, and beyond—alongside a deep commitment to open-source collaboration. He has played a key role in fostering community-driven innovation, having served as chairman of Tampere Entreprenourship society and co-founding Tampere Tribe to support local startup culture. Henri’s passion for AI, quality assurance, and rapid software development is evident in both his industry work and ongoing PhD research on agile product innovation. He frequently shares his expertise on stage and in publications, championing lean practices and the latest AI advances to empower organizations worldwide.

Highlights:

• AI systems amplify both mistakes and successes at scale, so the checks, guardrails, and validation processes built around the model matter more than the model itself.
• Testing AI requires a shift from deterministic pass/fail checks to monitoring trends and mean time between failures, because non-deterministic outputs cannot be verified with a single green test.
• The communication problem with AI agents is structurally identical to the bug-report problem with humans: vague input produces generic, context-free output that misses the actual need.
• As AI-generated code becomes a black box, test specifications and acceptance criteria become the primary source of truth, making the tester's skill set central rather than peripheral.
• AI democratizes software creation by removing the need for programming knowledge, which surfaces long-ignored organizational problems such as document version control and missing single sources of truth.