ИИ в британском бизнесе: внедрение утроилось за три года, но вглубь не идёт — ONS

The UK's Office for National Statistics (ONS) reported in July 2026: the share of British companies with 10 or more employees using artificial intelligence has grown from roughly one in eight three years ago to one in three today — almost a threefold increase.

How much has adoption grown?

AI adoption in British business has nearly tripled over three years: whereas around one company in eight (about 12%) used neural networks before, now it's one in three (about 33%), according to ONS data. Over this period, artificial intelligence has gone from a niche tool for specialists to a mass-market workplace technology.

• Three years ago — about 1 company in 8 used AI
• Now — 1 in 3
• Growth — almost threefold in three years
• Sample — British firms with 10 or more employees
• Data source — UK Office for National Statistics (ONS)

The "10 or more employees" threshold matters: this isn't just about tech startups, but about medium and large businesses for which AI has become an everyday part of the work. At the same time, "AI use" in the ONS survey is self-reported by the companies themselves, not an audit of how deeply the technology is embedded in their processes. Over three years, it wasn't just a handful of market leaders that raised the share of AI users, but the mass of mid-sized businesses.

Why is growth "wide, not deep"?

Most companies that have adopted AI use it superficially — for individual tasks, rather than as a technology embedded in core processes, notes The Next Web in its analysis of the ONS data. Breadth is expanding faster than depth: more and more firms are trying tools, but few are restructuring their workflows around them.

The difference between "adoption in breadth" and "adoption in depth" is the report's key storyline. A threefold increase in the number of companies that say they use AI doesn't mean the technology has become critical for them. For many, it's still experimentation: a customer-support chatbot, text generation, automating a single routine task. That kind of use easily shows up in the statistics but barely changes the business's economics.

The same ONS data also reveals the flip side of the growth: even when one company in three reports using AI, there's a large gap between "tried the tool" and "restructured processes around it." It's precisely in that gap that it gets decided whether the statistical growth turns into real productivity. While breadth keeps expanding, depth remains an area where most companies have made almost no progress.

That's exactly why the loud "tripling" figure can be misleading. It captures how many companies have at least touched AI, but says nothing about how many of them got a real effect from it — a productivity increase, cost savings, or new products.

"British companies have tripled their AI adoption.

But most are still barely scratching the surface," is how The Next Web frames the main takeaway.

What this means

The tripling statistic looks like a story of explosive success, but the real divide isn't between those who've adopted AI and those who haven't — it's between superficial and deep use. When adoption reaches a third of the market, the competitive edge shifts: the next phase for British business (and not only British) isn't "do you have AI" but "how embedded is it in your processes, and does it deliver a measurable result."