How can you tell if your team’s AI output is becoming generic?

Look for four signals: work that could carry a competitor’s name without anyone noticing, drafts across the team that share the same structure and rhythm, output that’s faster but no longer argued over, and writing with no detail only your business could have supplied. Any two together suggest drift.

Why this is hard to spot from the inside

The drift is invisible at the level of any single piece of work. Each draft looks fine on its own, often better than what the same person produced a year ago. The effect only shows up when you look across a team, or across time, and checking for it isn’t usually anybody’s actual job.

It also arrives quietly. There’s no single bad day when the work suddenly turns generic. It converges by small degrees, and every individual degree is perfectly defensible.

The four signals, in order of usefulness

1. The blind taste test. Take a recent piece of your team’s work and a comparable piece from a competitor, cover the names and logos, and ask someone outside the team to guess which is which. If they can’t tell, or they guess wrong, that’s your answer.

2. Structural sameness. Read 5 drafts from 5 different people in one sitting. Are they reaching for the same shape? The same 3-item lists, the same rhetorical opening, the same tidy summary at the end. Individual writers leave individual fingerprints, and shared AI tooling sands them off.

There’s a reason for this beyond coincidence. A 2024 study in Science Advances (Doshi and Hauser, 10(28): eadn5290, DOI: 10.1126/sciadv.adn5290) found that when writers used an AI-generated idea to start a piece of creative writing, their finished work came out 5.2% more similar to that specific starting idea than work produced without one, even though they were free to take it anywhere from there. The AI doesn’t just speed writers up. It anchors them to its own suggestion. Do that across a whole team using the same tool the same way, and the anchoring adds up to a house style nobody actually chose.

3. The disappearance of argument. In a healthy team, drafts get argued about. When AI-assisted work arrives fluent and already polished, it invites approval rather than debate, and the editing conversation quietly stops happening. Less friction feels like progress. It’s often the clearest early sign, because it’s the one that feels good while it’s happening.

4. Absence of the specific. Search a recent piece for anything only your organisation could have written: a named client, a real figure, an actual thing someone said, a detail from your own work. If it could have been written by someone who’s never set foot in your business, it probably was.

Two things that look like signals but aren’t

Faster turnaround. Speed on its own tells you nothing. Plenty of teams get faster and stay distinctive.

Team members using different tools. Untidy, but not the actual cause. The homogenising pull comes from how the tools are used, not which ones. A team all using one tool with skill and intention will produce more distinctive work than a team using 6 tools without any.

A simple thing to run this month

Take 10 pieces published across the last quarter, cover the names, and read them in one sitting. Count how many contain something only your business could have supplied.

The number is usually lower than you’d expect, and it’s a far more useful measure than any impression formed one piece at a time.

What to do if the answer is uncomfortable

The fix isn’t to stop using AI. It’s to change what the team puts into it and what they expect back: real briefs carrying real material, a shared standard everyone works to, and the editing conversation deliberately put back in.

That’s a training problem rather than a tooling problem, which is genuinely good news. Tools have to be bought again every year. Skill stays.

If you’re watching all of this happen and wondering how to make AI work for your business, book a call with me today – I train and consult businesses and teams to use AI with confidence, skill and intention.

Reference: Doshi, A.R. and Hauser, O.P. (2024). Generative AI enhances individual creativity but reduces the collective diversity of novel content. Science Advances, 10(28): eadn5290. DOI: 10.1126/sciadv.adn5290