Thoughts
Our thoughts, ideas, news and views that make us passionate about customer experience.
Our thoughts, ideas, news and views that make us passionate about customer experience.
Amazon has now blocked AI shopping agents from Perplexity, Google, OpenAI and Meta. Behind the legal arguments sits a UX problem: customers who ask an agent to buy for them and hit a wall. Here is what the standoff means for your customer journeys, and three questions to ask now.
Read the article: The shopping cart battleground: what Amazon's shopping agent standoff means for UXGenerative AI is changing how quickly teams design and build, but speed does not guarantee accessibility. Here are five things to consider, from putting accessibility into the prompt to testing the experience and knowing who stays responsible.
Read the article: Building Accessibility with Generative AI: Five Things Teams need to ConsiderWe're back on G-Cloud 15 with 76 specialist services spanning UX research, service design, accessibility and training - here's what's available and how it can support your 2027 planning.
Read the article: User Vision is on G-Cloud 15AI scribes save GPs time, but University of Edinburgh research suggests other things change too. Three lessons on behaviour, trust and effort for anyone introducing AI into a service people rely on.
Read the article: The Unintended Consequences of Making Things EasierTwo accessibility deadlines moved in opposite directions this year. The EAA turned one in June and enforcement is now visibly ramping up across EU member states, while the US extended its ADA Title II deadline for state and local governments to April 2027 or 2028, depending on size. Neither is a reason to relax: both frameworks converge on WCAG 2.1 AA as the baseline, and this piece sets out what the extra runway is actually for.
Read the article: The Deadlines Moved. The Standard Didn't.Vinted’s automated account bans may be intended to protect users, but when decisions are opaque, explanations are vague and appeals come only after the damage is done, the problem is more than technical. This article explores why AI-driven enforcement is fundamentally a process design issue—and what platforms can do to make automated decisions more transparent, fair and trustworthy.
Read the article: Banned by the Algorithm: Why "Appeal Later" Isn't Trustworthy AIEthical AI is not a values statement. It's a design specification. Here's what the EU's four core requirements actually demand of the people building AI products, and why most organisations are getting accountability wrong.
Read the article: Is Your AI Actually Trustworthy? What the EU Ethics Guidelines Mean for Design TeamsForbes' 2026 Accessibility 200 (published 19 May, 200 honourees, 23 countries) leans heavily on AI powered personalisation this year - from a tactile sign-language robotic hand to Amazon's AI shopping assistant. But 2026 also produced the first WCAG compliance regression in six years and a record year for accessibility lawsuits, a reminder that recognition for innovation and compliance are answering two different questions.
Read the article: Forbes Just Ranked Accessibility's Best. The Rest of the Web Got WorseThe ScotRail app problems are a reminder that a digital launch is only part of the story. This article explores why smooth transitions, realistic testing and accessible design are essential when people rely on a service under pressure.
Read the article: Designing the Journey to ChangeThe EU's Digital Omnibus just pushed the AI Act's high-risk obligations back 18 months. The accessibility clause buried inside them didn't get any easier to ignore — here's what actually changed, and what didn't.
Read the article: The Locked Front Door: Why the EU AI Act Makes Accessibility Non-Negotiable