Corporate AI strategy is no longer optional. Between 2024 and 2026, the share of companies making zero AI investment dropped from 15% to 4%, according to Designer Fund and Foundation Capital's AI in Design 2026 report. Figma's 2026 data identifies 27% of companies as 'Directive,' running top-down AI mandates, and those companies report collaboration impact nearly tripling compared to less structured adopters. The hype phase is over. ROI pressure, security failures, and workflow friction are forcing organizations to replace piecemeal tool-pushing with actual governance.
The gap between pressure and policy is still wide. N26 designer Marcus Knight built a Claude Code agent called Sheldon, trained to replicate the handoff questions of a senior engineer, with no company framework behind it, just a mandate to 'use AI somehow.' It worked. But Knight said it plainly: telling people to use AI is not a strategy. That friction is measurable. The AI in Design 2026 report found 73% of designers feel rising output pressure, while only 28% of surveyed leaders have made any formal process update to address it. Individual experimentation is producing real results. Institutional support is lagging.
The most extreme case is Fin, formerly Intercom, now acquired by Salesforce. Co-founder Des Traynor described the company's post-2022 pivot at UXDX EMEA in Berlin: strategy, structure, and name all rebuilt around AI after the product realized any SaaS feature could be cloned in days. The full account of that transformation, alongside the Figma adoption matrix, Knight's Sheldon experiment, and the closing gap between startup and enterprise AI support, is what makes this article worth reading in full. The numbers are directional. The case studies show the cost of waiting.
[READ ORIGINAL →]