Photo via Fast Company
When Anthropic announced Claude Design in mid-April, it sent shockwaves through the design software industry—particularly among its own partners. Figma and Adobe, both of which maintain multiyear collaborations with Anthropic to integrate Claude into their platforms, appeared caught off guard by the announcement. Figma's stock dropped 7% on the news, while Adobe declined 2.5%, reflecting investor concerns about potential market cannibalization. The surprise was especially notable because design tool companies rely on best-in-class AI models to remain competitive, yet they also depend on partnerships to maintain relevance in an increasingly AI-driven landscape.
The announcement exposed a growing tension in tech partnerships: the "frenemy" dynamic. Companies like Figma and Adobe need access to frontier AI models like Claude to power their products, but they also compete with the companies that build those models. Adding to the friction, Anthropic's chief product officer stepped down from Figma's board days before the Claude Design announcement, prompting Figma's cofounder to note that communications had not been consistently transparent. Only Canva appeared to have inside knowledge, claiming it co-developed Claude Design with Anthropic.
Despite the disruption, market data suggests established design platforms remain resilient. Adobe reported 11.5% revenue growth in 2025, while Figma achieved a striking 41% growth rate. Canva added 85 million new users and grew revenue 35%, indicating that designers value established tools' familiarity and power. According to Andy Allen, designer behind the early iPad app Paper, professional designers prioritize proven tools over experimental ones unless measurably superior—a pattern that has played out repeatedly as platforms shifted from Quark to InDesign, Freehand to Illustrator, and Sketch to Figma.
The broader competitive dynamic reveals that design tool companies are positioning themselves as AI "curators" rather than singular platforms. Figma and Adobe both emphasize model-agnostic approaches, integrating multiple AI systems where each excels. This strategy insulates them from over-dependence on any single frontier model, since swapping AI engines requires just a line of code change. Meanwhile, companies like Anthropic face their own vulnerability: training and operating large language models remains prohibitively expensive, and there's no guarantee design platforms will maintain Claude integrations if competing models prove superior.
