Can We Trust AI Detection? Inside Pangram Labs

Can We Trust AI Detection? Inside Pangram Labs

In this episode of A Journalist’s Guide to AI, I sit down with Max Spero, co-founder of Pangram Labs, to explore one of the most important questions in AI right now: what happens when we start relying on it a little too much?

This conversation goes beyond just whether we can detect AI-generated content. We dive into why it matters, for how we learn, how we communicate, and ultimately, how we trust one another.

Max describes AI as a completely new paradigm, and one idea that really stuck with me is this: when we overuse AI, we may not actually be learning as effectively as we think we are.

There are countless AI tools being built every day, but Max shares that his motivation for starting Pangram came from a deeper concern, the future of communication itself. We even explore the possibility of a world where AI agents mostly talk to other AI agents, and why, as Max puts it, that could be a pretty grim outcome.

We also take a closer look at AI detection tools, why so many of them fall short, and the real risks of false positives, especially in newsrooms and classrooms. This isn’t just a technical issue; it’s one that can have serious human consequences.

What I really appreciate about this conversation is the focus on balance. Verification isn’t about punishment—it’s about accountability, trust, and making sure humans stay at the centre of this evolving landscape.