Across multiple recent studies, the same pattern keeps showing up: even when AI-generated content is polished, people react to it more warily once they know a machine was involved. Work identified as AI-made gets rated as more generic, more forgettable, and less trustworthy than equivalent work attributed to a human — even when the underlying content is identical.
The effect isn’t subtle, either. In one recent industry survey on AI disclosure, roughly twice as many people said learning that content was AI-generated made them trust a brand less as said it made them trust the brand more. That’s a meaningful skew, and it points to something important: the issue usually isn’t the AI-generated work itself. It’s what people assume that work implies about everything else the business is doing.
This is where your instinct about “what other shortcuts are they taking” comes in. Distrust of AI doesn’t stay contained to the AI-touched part of a project. It generalizes. Broader research on public attitudes toward AI in business and institutions consistently finds that a large majority of people don’t trust organizations to use AI responsibly in the first place. Once that baseline skepticism exists, a visibly AI-built website doesn’t just look like a website — it looks like evidence.
Why AI-built sites actually do tend to feel generic
Here’s the part that’s easy to miss: the “cheap and generic” perception isn’t just an image problem — it’s often an accurate read.
When a generative tool is given a vague brief — something like “build a clean landing page” — it defaults to the statistical average of every landing page it’s seen. No strong point of view, no product-specific direction, nothing distinct. The result is generic, because generic is exactly what an underspecified prompt produces. The tool isn’t failing; it’s doing precisely what it was asked to do, which was nothing in particular.
That’s a useful reframe. The problem was never “AI can’t make good design.” The problem is design without direction — and AI just makes it faster to ship undirected work at scale, which is exactly what erodes client and end-user trust.
What the design industry is doing about it
The interesting shift happening right now isn’t “avoid AI.” Most working designers have already integrated it into research, layout exploration, and production. The shift is toward treating visible craft as the counter-signal.
As AI-generated visuals become more common, deliberately human touches — organic shapes, asymmetry, texture, custom illustration, imperfection — are increasingly what signal warmth and trust in an otherwise AI-saturated landscape. The logic is straightforward: when everything can be generated instantly, restraint and intentional design decisions become the thing that can’t be faked cheaply. People may not be able to articulate why a site feels handcrafted versus templated, but they register the difference anyway, and it shapes how much they trust the business behind it.
That’s the opposite of what a lot of AI-website messaging promises. The pitch is usually speed and cost — a site in minutes, no designer required. But speed and cost are exactly the signals that trigger the “what did they cut corners on” reaction. The businesses actually winning with AI in their workflow aren’t the ones broadcasting how fast they built the site. They’re the ones using AI invisibly, in service of a stronger, more specific creative direction, and letting the craft do the talking.
The takeaway
If you’re a business owner evaluating a web design partner right now, the AI question isn’t really “do they use AI or not” — almost everyone does, at some stage. The better question is whether the tool served a clear creative point of view, or replaced the need for one. The first produces work that still gets scrutinized and holds up. The second produces the kind of site people can spot from across the room — and once they’ve spotted it, they start wondering what else got the same treatment.
