The AI landscape doesn't move in one direction — it lurches. Some techniques leap from experiment to table stakes in a single quarter; others stall against regulatory walls, technical ceilings, or organisational inertia that no amount of hype can dislodge. Knowing which is which is the hard part. The State of Play cuts through the noise with a rigorously maintained index of AI techniques across every major business domain — classified by maturity, evidenced by real-world adoption, and updated daily so you always know where you stand relative to the field. Stop guessing. Start knowing.
A daily newsletter distilling the past two weeks of movement in a domain or two — delivered to your inbox while the index updates in the background.
Each dot marks the weighted maturity of practices within a domain — hover for a brief summary, click for more detail
AI applied from user research through to shipped product experience. Wide maturity spread: A/B testing and analytics are established, prototyping and design systems are good practice, but nearly half the domain is bleeding-edge — generative UI, autonomous UX research, and AI-native product frameworks are experimental. Most practices are stalled, with more energy in tooling announcements than production adoption.
The headline: AI can now design, write, prototype, and analyze your product faster than your team can — but only a small group of disciplined organizations are getting real returns, and the gap between them and everyone else is widening fast.
The tools have stopped being the problem. AI design and product software is now genuinely capable: it generates working code from a sketch, writes on-brand copy, builds research summaries in hours instead of weeks, and runs autonomous agents (software that acts on its own without being prompted) that spot problems in your usage data. Adoption is near-universal — most designers now use AI every week, and half ship AI-written code to production. But capability and results have split apart. A small vanguard of well-run companies is pulling clear returns: faster development, higher margins, real cost savings. The rest are stuck — roughly 95 percent of organizations report no measurable financial return from their AI spending. The difference is almost never the tool. It is whether the organization built the rules, standards, and review process to use it safely. The window to be in the first group, not the second, is the thing to watch.
The big design vendors made their tools work better together. Google, Anthropic, and Apple all shipped infrastructure this fortnight that lets AI follow your brand's design rules automatically instead of inventing its own. This matters because the companies getting value share one trait — they encode their standards as machine-readable rules first, then let AI work inside them. If your design system is just documentation, it is not ready for this.
Fresh data confirmed AI raises output and defects at the same time. A study of 22,000 developers found AI lifted task completion by a third — but production incidents jumped sharply, and senior engineers now lose up to a third of their week cleaning up AI-generated work. The lesson is not to slow down adoption; it is to budget for the review and cleanup that the headline speed numbers hide.
Accessibility resurfaced as a legal and quality landmine. AI-built interfaces are accessible by default only two-thirds of the time, and one major vendor's new design-to-web tool shipped demo content with over 200 accessibility violations. With more than 5,000 disability-access lawsuits filed last year, untested AI-generated interfaces are now a direct liability — not a niche compliance concern.
Otherwise, the landscape held steady. No practice in this domain moved up or down this fortnight. The story is consolidation among leaders, not a breakthrough that changes the playing field.
Accessibility deadlines are live and enforcement is real. US Department of Justice deadlines for public-sector digital accessibility run through 2027–2028, and the legal record now shows that automated "overlay" widgets provide no protection while documented source-code fixes win cases. Confirm someone owns accessibility for anything AI helped build, and that fixes happen in the actual code.
Regulators are circling personalization and tracking. Privacy lawsuits over session-recording tools (800-plus filed last year), a Texas probe into Spotify's pay-to-promote feature, and EU transparency rules are all tightening. If you personalize experiences or record user sessions, get legal and product aligned on disclosure and consent in the next two quarters.
The skill gap, not the tool gap, will decide who wins. As AI makes building cheap, the scarce skill becomes writing clear requirements up front — vague instructions now cost more because endless rework feels free. Invest in specification discipline and review capacity, not just more AI seats.
The same tool produces opposite results depending on your discipline. Organizations with built-in governance report 18 percent higher margins and run many times more AI projects successfully; those without get nothing from identical software. Buying the tool is the easy 10 percent of the work.
You cannot trust the output without checking it, and checking is the new bottleneck. AI confidently makes things up (hallucination) at rates that better models have not fixed — research summaries built on fabricated findings, usability "problems" that do not exist, analytics answers that are wrong but persuasive. Teams now spend a large share of their time verifying AI output, which quietly eats the speed it promised.
Adoption is easy; quality is not. Your team will love the speed and report glowing satisfaction. The defects, incidents, and brand drift show up later and somewhere else — in production, in legal, in customer trust. Measure success on business outcomes, not on how many people use the tool.
Go deeper: the full Product & Design briefing — the longer analytical write-up, plus every practice we track in this domain with its maturity rating, the tools to consider, and the evidence behind our assessment.