UK firms drown in AI hype, emerge with empty spreadsheets

Nearly 80% of UK firms now use AI tools, yet barely any report a positive return on investment. The numbers, sourced from TechRadar, reveal a stark disconnect: adoption is booming, but measurable outcomes are nowhere to be found. Weak planning, unclear goals, and inconsistent tracking are the usual suspects—hardly surprising, but increasingly indefensible as AI moves from novelty to necessity.

The real question isn’t whether companies are using AI, but why they’re using it. Early signals suggest many firms are deploying low-impact tools—chatbots, basic automation—while calling it a ‘transformation.’ McKinsey’s latest AI report notes that high-impact deployments (predictive analytics, generative AI) remain rare, often confined to tech giants with the resources to experiment. For everyone else, AI adoption looks less like strategy and more like FOMO with a budget.

This isn’t just a UK problem. The same pattern plays out in the US and Europe, where Gartner’s 2023 AI survey found that 54% of organizations are piloting AI, but only 14% have scaled it. The gap between demo and deployment remains vast, and the marketing teams aren’t helping. Vendors sell ‘AI-powered’ solutions like they’re plug-and-play, while businesses scramble to justify the spend.

The developer community isn’t buying the hype. GitHub activity around enterprise AI tools shows steady but cautious engagement—lots of forks, few stars, and even fewer production commits. Hacker News threads on AI ROI are filled with war stories: ‘We built a chatbot. It cost £200k. It answers three questions.’ The frustration is palpable, but so is the resignation. AI is now table stakes, even if the table hasn’t been set yet.

So who wins here? The vendors, for now. Cloud providers, AI startups, and consultancies are the only ones seeing consistent returns, while businesses foot the bill for experiments that don’t scale. The competitive advantage isn’t in adopting AI—it’s in adopting it well. That means clear metrics, structured strategies, and a willingness to kill projects that don’t deliver. Until then, the 80% adoption rate is less a sign of progress and more a measure of how effectively the AI industry has sold the dream.

The real signal isn’t in the adoption numbers, but in the silence around ROI. If nearly everyone is using AI and almost no one is profiting from it, the problem isn’t the technology. It’s the execution—or the lack thereof.