Tell people you used AI and they trust you less. That instinct is quietly turning human-made into the signal.

You add a small, honest line to the bottom of the report, the email, the post: drafted with AI assistance. You expect a little credit for being transparent. Instead the response cools. The work is exactly the same as it would have been; the disclosure is the only thing that changed, and it changed how the work landed. That is not paranoia or a one-off. It is a measured effect, and it points at something larger than a communications quirk.

In a 2025 paper in Organizational Behavior and Human Decision Processes, researchers ran a series of controlled experiments testing a simple question: does disclosing that you used AI compromise trust in you? Across the studies, and in a combined analysis of them, it did. Disclosing AI use reduced how much people trusted the discloser. And the effect was stubborn. It held even when the person clarified that a human had reviewed and revised the work, and even when they specified that AI had only been used for proofreading. Transparency, the very thing we are told builds trust, quietly cost it.

Why would being honest about AI reduce trust?

Because trust runs on legitimacy, and legitimacy runs on norms. We still carry a mostly unexamined assumption that this kind of work is done by a person, so disclosing AI signals a departure from that norm, and the mind reads the departure as a red flag. The study traces the erosion through what people infer from the disclosure: the work seems less typical, the person seems less committed, having apparently put less of themselves into it, and the whole thing feels less authentic. Notice what is absent from that list: quality. The trust drop is not a judgement about whether the work is good. It is a judgement about how much genuine human effort, ownership and care sits behind it. That is a question about what we perceive as real, and it sits close to the centre of how humans decide what to believe.

So should you just hide it?

No, and this is exactly the trap. The research frames it as a dilemma precisely because the alternative is worse. Undisclosed AI use that later surfaces reads as concealment, and concealment breaks trust far more decisively than disclosure ever dents it. You are choosing between a small, honest cost now and a large, dishonest one later, which is not really a choice at all. The way out is not to hide the tool. It is to change what you attach to it.

What actually restores the trust?

Put back the thing the bare disclosure stripped out: visible human commitment and accountability. The penalty comes from the sense that no one truly stood behind the work, so make it unmistakable that someone did. "AI helped produce this, I have read every line, and I stand behind it" carries a legitimacy that "produced with AI" removes, because it re-supplies the human judgment and ownership the disclosure appeared to remove. You are not concealing the machine. You are making the person visible again. Done honestly, that is not spin; it is an accurate description of who is responsible.

1. Do not hide AI use. Concealment is the bigger trust risk. Disclose, but never as a bare "made with AI" tag.
2. Pair disclosure with human accountability. Name the person who exercised judgment and stands behind the outcome. That ownership is what restores legitimacy.
3. Protect the human layer where trust is the product. In advice, brand, relationships and care, keep genuine human judgment in the loop, not as theatre but because it is what people are actually trusting.
4. Treat human-made as an asset, not an embarrassment. As synthetic content floods in, verifiable human effort is becoming a differentiator. Where your work is genuinely human, make that legible.
5. Never fake the human signal. Claiming effort or oversight that was not there is the concealment trap in reverse. It collapses the moment it is tested.

What does this really tell us?

The finding looks like a quirk of communication. It is really a window into how humans decide what is real and what to trust. Mostly without noticing, we treat human effort as evidence of commitment, care and accountability, and we treat its absence as a reason to trust less. That instinct will not fade as AI improves. If anything it sharpens as synthetic content becomes both indistinguishable and everywhere, because when you can no longer tell from the surface, provenance is all you have left to go on.

That is the paradox leaders now sit inside. Transparency about AI costs you in the short run, while the very instinct behind that cost is what makes protected, visible, genuine human judgment more valuable over time. When anything can be AI-generated, human-made stops being the default and becomes the signal, not out of nostalgia, but because it is what people are actually reaching for when they decide whom to believe. That is the subtler edge of the end of business as usual: the scarce, trusted thing is no longer the content, which is now infinite, but the human willing to stand behind it. Protect that, make it legible, and never fake it, because it is quietly becoming the most valuable part of what you offer, in line with the truth that your advantage is what is distinctly yours, not the tools everyone can buy.

About the author

Thomas Green

British technology futurist, AI keynote speaker and advisor. Thirty years across enterprise technology and AI strategy, helping leaders navigate the future of work. The futurist who died.