Ten years of responsible AI work has produced real practice for spotting and reducing harm in high-stakes systems. What it has not produced is a market that pays for any of it. A company that invests seriously in safety, fairness and oversight has no reliable way to prove to customers, regulators or shareholders that it did more than the legal minimum.

That is a structural problem rather than a moral one. When the careful firm and the box-ticking firm look identical from the outside, spending on care is a competitive disadvantage. The authors call the resulting condition a trust gap, and their argument is that it persists because society lacks a way to recognise or compare the difference.

Their proposed answer is independent certification. The precedent is familiar from other industries: a restaurant hygiene rating in the window works not because inspectors are infallible but because it makes an invisible quality visible at the point of choosing, which is what lets the market reward it at all.