If scrolling through a latest app feels familiar, the design may have been generated by artificial intelligence.
The beige blur of AI-coded apps
Tools like Claude Code, Lovable, Replit, and Base44 allow non-technical founders to build apps in hours. Many share a similar look, described by researchers as “regression to the mean.”
Paul Bakaus, CEO of AI design startup Impeccable, compared it to an “algorithmic Uniqlo or Ikea”—functional but not unique. Donghoon Shin, a human-computer interaction researcher at the University of Washington, found these apps often use muted grays and whites, a single brand color, and rounded corners with drop shadows. Sauvik Das, an associate professor at Carnegie Mellon, said the effect is so widespread that users now call it “AI slop.”
Priyanshi Bansal, a product manager in India, created an AI-generated gift-picking app that early users called generic. “The first version had too many emojis, shadows, and rounded edges,” she said. Bansal is now redesigning it with a cleaner interface.
Related: Family creates core memory at Mets game
Polished surfaces, half-baked functions
A second issue is the gap between appearance and usability. AI tools often produce sleek landing pages for apps still in early development, Das explained. The designs look finished until someone tries to use them.
Ankush Samant, a lecturer at the National University of Singapore, noted that human designers consider button weight, onboarding pacing, and error messaging—details AI misses. “AI focuses on the ideal scenario,” he said. “It might make a button look clickable, but the function won’t work.”
Shin pointed out that AI rarely handles edge cases like empty states, skeleton loaders, or offline modes. Error messages, when included, are often vague placeholders. Samant said these moments show whether an app was built for real users or just to impress in a demo.
This creates real problems. Apps that succeed commercially need strong UX, not just visual appeal.
Related: Coffee chain targets undercaffeinated US regions
The tools continue to improve. Samant suggested changing prompts from aesthetics to user emotions. Shin recommended giving AI specific brand guidelines and examples of what not to do.
Das said product-market fit matters more than perfect design, but a professional designer still plays a key role in refining an app. “AI helps iterate on interfaces,” he said. “But creating one from scratch remains a human task.”
The common signs include a neutral background, one accent color, and buttons that don’t deliver what they promise. For now, these patterns make AI-generated apps easy to spot.
Leave a Reply