This week’s design and UX headlines converge on a single theme: the gap between looking competent and being competent is widening fast. Whether it’s AI-generated software, accessibility compliance, or design leadership, the industry is calling out surface-level execution. Here’s what our team is tracking and how we’re applying these insights across client projects right now.

Key Takeaways

• AI won’t kill software — it will force it to evolve. Websites built on rigid, conventional logic face the steepest decline.
• 95.9% of the top one million websites still fail basic accessibility checks. Designers already have the skills to fix this.
• Starting with AI tools is easy. Getting reliable, quality output from them requires structured oversight and domain expertise.
• Design leadership needs to shift from mechanical process management to adaptive, judgment-led decision-making.
• Psychological ownership — feeling responsible for outcomes — boosts effectiveness, but only when directed at things you can actually control.

Software Isn’t Dying — It’s Being Forced to Adapt

Enterprise software stocks took a hit in August 2025, and the knee-jerk reaction was to blame AI. But as explored in this analysis on the future of software in an AI-saturated market, extinction is the wrong lens. Evolution is the right one. AI models are getting better at building and doing, which means conventional, static software faces pressure — but adaptive, well-designed digital products will thrive.

For our clients, this means websites can no longer be static brochures. We’re building sites that integrate dynamic content, personalisation layers, and AI-assisted functionality where it genuinely serves the user. The bar for what counts as a “good” website just moved up.

Accessibility Isn’t a Separate Skill — You Already Know How

WebAIM’s 2026 crawl found errors on 95.9% of the top one million homepages, averaging multiple failures per page. That number is staggering, but as argued in this practical guide for designers who think accessibility is a separate skill set, the fix isn’t learning something new. It’s applying what you already know — proper contrast ratios, logical heading structures, descriptive link text, keyboard navigation.

We treat accessibility as a core design deliverable, not an afterthought audit. Every site we build goes through structured checks against WCAG standards before launch. The data proves most agencies skip this. We don’t.

AI Adoption Without Oversight Creates Real Damage

There’s a dangerous confidence gap emerging. AI tools look reliable before they actually are. This breakdown of the gap between AI reliability and AI appearance puts it bluntly: getting started is not getting it right. Three years of building with AI — from UX writing plug-ins to full agent workflows — reveals that unstructured adoption leads to inconsistent quality and real business risk.

Our approach: we use AI to accelerate specific, well-defined tasks within our design and development workflows. Content drafting, code scaffolding, image optimisation. But every output gets human review. No exceptions. The moment you trust AI output without verification is the moment quality drops.

Design Leadership Must Move Beyond Mechanical Thinking

Mechanical systems are brittle. That’s the core argument in this week’s curated resources on rethinking leadership and judgment as a bottleneck. Design leaders who fixate solely on roadmaps, cross-functional alignment, and operational process miss the point. Judgment — the ability to make good calls under uncertainty — is the actual bottleneck.

We’ve restructured our project leadership around this principle. Senior designers own decisions, not just deliverables. That means faster iteration, fewer approval loops, and better outcomes for clients.

Direct Your Ownership Where It Counts

Psychological ownership — the feeling that something is “yours” — is powerful. But as Nielsen Norman Group’s research on psychological ownership makes clear, it backfires when attached to things outside your control. Designers who feel ownership over a pixel-perfect comp but not over user outcomes are optimising the wrong thing.

We coach our team to own results — conversion rates, page speed scores, accessibility compliance — not just visual output. That shift changes how you design from the ground up.

The common thread across all five of these developments is clear: surface-level competence is no longer enough. Whether you’re adopting AI, leading a design team, or building a website, the gap between looking good and performing well is where businesses lose money. Our job is to close that gap for every client we work with.

Frequently Asked Questions

What does AI mean for the future of website design?

AI won’t replace well-designed websites — it raises the standard. Sites need to be more adaptive, accessible, and performance-driven to stay competitive against AI-powered alternatives.

How do web designers make a website accessible without specialist training?

Core accessibility fixes — proper contrast, logical headings, descriptive alt text, keyboard navigation — use skills every competent designer already has. The issue is prioritisation, not knowledge.

Why does using AI in web design still require human oversight?

AI output often looks polished but contains factual errors, inconsistent logic, or accessibility failures. Human review catches these issues before they reach a live site and affect real users.

What is psychological ownership in UX design?

It’s the feeling of personal responsibility for a project or outcome. It drives better work when directed at measurable results like conversions and usability, but causes friction when attached to elements outside your control.