A product lead at a mid-size SaaS company recently showed us her team's new setup wizard, built almost entirely with an AI design tool: a progress bar across the top, a card-based step layout, a friendly illustration on the right, a "skip for now" link in the same gray as every other skip link on the internet. It was clean. It was also, by her own admission, indistinguishable from three competitors' onboarding flows she pulled up next to it. The same pattern shows up everywhere Aero works: a biotech patient portal whose dashboard cards could be swapped with a fintech app's without anyone noticing, a VC's deal-flow tool that reads like a template, a media brand's article page that lost the one visual habit longtime readers used to recognize it. None of these teams did anything wrong. They used a fast, capable tool the way it was designed to be used. The result is a quieter problem: products that work fine and disappear into each other.
AI design tools are no longer a novelty. In Figma's 2026 State of the Designer report, 91 percent of designers surveyed said new AI tools improve their designs and 89 percent said the tools help them work faster, according to Figma's own research. That is a fast, broad shift in how interfaces get built, across industries and team sizes. It is also exactly the mechanism behind the sameness problem. These tools generate and suggest patterns trained on what already works well across many products: the most common layout, the most common component, the most common microcopy. That is a feature when a team needs a working first draft fast. It becomes a liability when every team accepts the same default, because the tool has no reason to know or protect what makes any single product recognizably itself.
No individual decision to accept an AI-suggested layout, spacing value, or component feels like a brand decision. It feels like shipping faster. But brand differentiation is rarely lost in one dramatic redesign. It is lost in hundreds of small defaults accepted because they were fast and looked fine in isolation. A design system with no opinion about which of those defaults matter will absorb whatever the AI tool suggests, one screen at a time, until the product's visual identity is whatever the model's training data considered typical.
Two B2B SaaS companies in adjacent markets rebuilt their onboarding flows within a few months of each other, both leaning heavily on AI design tooling to move fast. The first team accepted the tool's defaults across the board: layout, component styling, even most of the copy tone. Their new flow tested fine for usability, but internal reviewers and a handful of prospects independently described it as "generic" or said it reminded them of a competitor's product. The second team used the same class of tool for scaffolding, structure, and first-pass copy, but ran every screen through two brand-specific filters before shipping: does this still use our system's distinct interaction pattern for progressive disclosure, and does the copy sound like something only this product would say. The scaffolding time savings were nearly identical between the two teams. The second team's product still felt like itself when it shipped. The difference was not the tool. It was whether a human decision point existed between "the AI suggested this" and "this ships."
This is Aero's own practical lens for catching sameness before it ships, not an established industry framework. Five questions we run against any interface built with heavy AI design tooling:
A team that answers "no" or "not sure" to two or more of these is likely shipping AI-assisted work faster than it is protecting what makes that work theirs.
The fix is not slowing down AI-assisted design. It is deciding, at the design system level, which patterns are non-negotiable brand signatures and which are safe to let an AI tool generate freely. We covered the related discipline of making a design system ready for AI-assisted and agentic work in is your design system agent ready, and the human role that has to sit inside that process in the design system engineer as human in the loop. Differentiation is the piece those two pieces leave implicit: a design system tells an AI tool what is allowed to vary and what is not. Without that instruction, the tool defaults to the most common answer, and the most common answer is, by definition, not what makes a product distinct.
No. The speed gains are real and worth keeping. The fix is adding a deliberate checkpoint, using the five questions above or something like them, rather than assuming speed and differentiation take care of themselves together.
Often, yes. Familiar patterns for navigation, forms, and system feedback reduce cognitive load and should usually stay familiar. The risk is treating every pattern, including the ones that carry your brand's identity and voice, as equally safe to standardize away.
Immediately. Pick your most visible surface, the one prospects or investors see first, and run the audit above against it before the next AI-assisted redesign ships. It takes an afternoon and surfaces exactly where defaults have crept in.
Worried your product is starting to look like everyone else's? Talk to Aero.
AI design tools now touch most product teams' workflows, and that convenience carries a real risk: interfaces that quietly converge on the same patterns. Here is Aero's practical framework for using AI design tools without losing what makes a product recognizably yours.
AI personalization is now standard across SaaS, fintech, and healthcare products, but accuracy alone doesn't earn user trust. Here is Aero's practical framework for designing personalization that discloses what it's doing instead of quietly guessing.
Most AI features are judged on accuracy, but users decide whether to trust an output long before they can verify it is correct. Here is Aero's practical framework for designing AI explainability, and why that is a different problem than the disclosure rules already on the books.
