Foundation Capital: how AI is reshaping design in tech

Foundation Capital and Designer Fund published this research jointly in July 2026, drawing on surveys of more than 900 designers at startups, enterprises, and agencies. The data was supplemented with interviews at teams from Anthropic, Framer, Linear, Notion, Shopify, Sierra, and Stripe, making it one of the broader empirical snapshots of what design work looks like in a year when AI tools have moved from experimentation to daily use.

The headline finding is a sharp shift in practice. Three-quarters of designers now use AI every day, and 91% use it for design tasks at least weekly, up from 53% the previous year. The change is not confined to a subset of technically inclined designers. Half of respondents say they have shipped AI-generated code to production, despite only 20% identifying as design engineers. The practical boundary between design and development has become functionally less meaningful for a significant portion of working designers.

Toolstack expansion reflects the pace of change. The average designer went from three tools to seven in a single year. Claude jumped from 52% to 78% adoption, surpassing ChatGPT to become the most widely used AI tool in this cohort. Figma is described as shifting from a final destination to a thinking surface, with prototypes increasingly replacing static mockups as the primary design output in some teams.

The research does not frame this as a replacement story. Eighty percent of designers say human judgment remains essential for creative direction and quality decisions. The consistent pattern across case studies is that AI handles generation and iteration while designers retain responsibility for strategy, critique, and final calls. Sixty-five percent of respondents report taking on more work that previously belonged to product management or engineering — a role expansion rather than a contraction.

The most persistent challenge identified is output quality. Sixty-two percent cite inconsistent or unreliable AI output as their biggest obstacle, pointing to a skill gap in prompt design and output evaluation rather than a fundamental limitation of the tools.

The report is most useful for design team leads and hiring managers assessing what the role expects in 2026, and for individual designers who want to benchmark their own AI adoption against what is currently typical in the field.