Half of UK Businesses Now Use AI. But Only 10% Go Beyond ChatGPT.

Four days ago, the British Chambers of Commerce published the most comprehensive study yet on AI adoption in UK businesses. The headline number got attention: 54% of UK firms are now actively using AI.

That's up from 35% last year. 25% the year before that. And 23% the year before that.

But the number that matters is not 54%.

It's 10%.

The two-speed AI economy

The BCC study — conducted with Atos and the University of Essex across 668 businesses — found something the headline missed. There are now two very different groups of businesses using AI.

Group 1: Generic AI users (roughly 44% of businesses). These firms use ChatGPT, Copilot, or similar off-the-shelf tools. They generate emails, summarise documents, brainstorm ideas. It's helpful. It saves some time. But it's fundamentally the same as what every competitor can do by opening the same app.

Group 2: Bespoke AI users (roughly 10% of businesses). These firms have AI integrated into their actual workflows. Custom systems connected to their CRM, their accounting software, their client management tools. The AI doesn't just answer questions — it does work. Automatically. In the background. While they focus on clients.

Here's the gap that should worry you: bespoke AI users are significantly more likely to report productivity gains and are more likely to be restructuring their operations around what AI makes possible. The generic users report modest benefits but haven't fundamentally changed how they work.

The numbers behind the shift

The BCC report breaks it down:

• 54% of UK firms now use AI (up from 35% in 2025)
• 95% of those firms say AI has had no impact on headcount — AI is augmenting, not replacing
• 86% say job roles are unchanged — people do the same jobs, just with better tools
• +71 percentage point net productivity improvement among current AI users
• Larger SMEs and B2B professional services lead adoption — smaller firms and consumer-facing businesses lag behind
• Only 14% of firms investing in AI training expect headcount reductions — the rest expect to grow

The picture is clear. AI adoption is accelerating, but it's deeply uneven. The firms that are winning are not the ones with the biggest budgets. They're the ones that have moved from "using AI" to "building with AI."

What "bespoke AI" actually looks like

Let me make this concrete. I build AI systems for UK service businesses, so here's what the 10% are actually doing differently.

A 15-person accountancy firm has an AI system that monitors their Xero accounts overnight. It reconciles bank transactions automatically, flags anomalies, and drafts client emails about outstanding invoices. The partners review the output over coffee. Previously, a junior spent two days a week on this.

A recruitment agency has an AI agent that screens incoming CVs against job specifications, scores candidates, and generates shortlist reports with reasoning. The recruiter still makes the final call, but instead of spending 3.6 hours per vacancy on initial screening, they spend 20 minutes reviewing the AI's recommendations.

A marketing consultancy has a content pipeline where AI researches industry trends, drafts social media posts in the client's brand voice, and schedules them across platforms. The account manager approves a week's content in 15 minutes instead of spending half a day creating it.

None of these required the business owner to learn to code. None required a six-figure investment. Each was built and running within two weeks.

Why the gap will widen

The BCC report recommends the government establish "AI champions" to help SMEs adopt AI, along with tax credits and training subsidies. The AI Minister said they're "helping 10 million people build practical AI skills."

That's encouraging. But here's the reality: government programmes take time. Your competitors aren't waiting.

The productivity data from the report tells the story. Firms already using AI expect significant productivity improvements (+71pp net). Firms still planning to adopt? Their optimism is "far lower." The gap between the two groups isn't static. It compounds.

Every month a competitor's AI handles their admin while yours is done manually, they're reinvesting that time into growth, clients, and service quality. After a year, the difference is visible. After two, it's structural.

The honest truth about getting started

Here's what I tell every business owner who asks me about AI:

You don't need to understand how it works. You need to understand what it should do. "Check my CRM for overdue invoices every morning and draft follow-up emails" is a clear enough brief. The technical implementation is someone else's job.

Start with one process, not a strategy. The firms in the BCC study that report the highest productivity gains started small. One workflow. One pain point. One system that saves 5+ hours a week. Then they expanded.

Generic AI tools are a starting point, not a destination. ChatGPT is useful. But it's the same tool available to every competitor. The competitive advantage comes from AI connected to your specific data, your specific workflows, your specific clients.

The security question has been answered. This was the biggest objection I heard from regulated industries throughout 2025. "What about client data?" With recent developments in agent security (deny-by-default sandboxing, auditable policy files), the answer is now a technical one, not a philosophical one. The tools exist.

What this means for you

If you're in the 46% not yet using AI — the window is closing. Not because AI will replace you next month, but because the 54% who are using it are pulling ahead on productivity, client service, and capacity to grow without proportionally growing headcount.

If you're in the 44% using generic AI — you've taken the first step. The question now is whether you stay in the "same tools as everyone else" group or move into the 10% building AI into your actual operations.

If you're already in the 10% — you know the advantage. The question is what's next.

The BCC data is clear: AI has moved from the margins to the mainstream. The businesses that treat it as a strategic capability rather than a novelty tool will define the next decade of UK commerce.

Sources

British Chambers of Commerce, "Future of Work: AI in the Workplace Report" (18 March 2026). Research conducted with Atos and University of Essex ESRC Centre for Micro-Social Change (MiSoC). 668 businesses surveyed, 94% SMEs. Fieldwork: 12 January — 6 February 2026.