UK data shows a jump in business adoption of AI technologies

The UK Office for National Statistics reported that nearly three in ten businesses now use at least one artificial intelligence technology. The figure does not prove a full productivity transformation, but it shows that AI is moving beyond isolated experiments and into broader business routines.

What happened

The Business insights and impact on the UK economy bulletin, published July 2 by the Office for National Statistics, says 29% of UK businesses used at least one type of AI technology in June 2026. That is up 8 percentage points from June 2025 and continues an upward trend over the last three years.

The increase is stronger among large companies. For businesses with 250 or more employees, the share reached 49%, up 13 percentage points from a year earlier. The most common use was text generation with large language models, reported by 17% of businesses. Visual content creation followed at 14%. GOV.UK published the same statistical release as part of the British government’s official statistics calendar.

Why it matters

The editorial value of the data is its restraint. This is not a vendor promise or a lab demonstration. It is an official survey of businesses. That makes it useful for observing adoption with less commercial noise: how many organizations say they use AI, which kinds of tools are entering first and where there is still distance between basic use and deeper transformation.

The most important reading is that adoption is growing, but not necessarily evenly. Large businesses appear to have more capacity to test, integrate and govern AI tools. Small and midsize firms may face different barriers: cost, limited internal talent, data concerns, difficulty measuring return or fear of depending on outside platforms.

What changes for users, companies and the AI ecosystem

For workers, the figure suggests that generative AI is becoming part of more workplaces, especially in drafting, summarizing, document analysis, visual ideation and administrative support. That can raise productivity, but it also requires clear rules for confidentiality, human review and output quality.

For companies, the signal is practical: if AI use is approaching one in three organizations in a developed market, the question is no longer whether to pay attention. It is how to adopt the technology with control. Adoption without policies can create errors, data exposure or dependence on tools that nobody audits. Moving too slowly can leave internal processes behind more organized competitors.

Social and strategic context

Lía Torres’ reading is that these numbers bring the AI conversation closer to the ground. Not every business is replacing workers with autonomous agents, and not every company is capturing massive productivity gains. What is verifiable is more concrete: more firms are incorporating some form of AI, and the main entry point remains content generated by language models.

That nuance matters because public policy and business planning must distinguish between use, dependence and transformation. Using a chatbot to draft text is not the same as redesigning an entire workflow. But if millions of small tasks change, the cumulative effect on skills, training, service quality and internal organization can be significant.

What remains unclear

The ONS warns that these are official statistics in development and advises caution when using the data. The bulletin does not by itself measure productivity gains, decision quality, job substitution or financial return. It also does not fully explain whether companies are using public tools, internal systems or AI features embedded in enterprise software. The data should be read as a sign of rising adoption, not proof of definitive economic impact.

Sources consulted

Office for National Statistics: Read More Read More by Lía Torres — Social and strategic perspective.