Only 21% of businesses are getting ROI from AI, here's why
TL;DR
A DataCamp survey of over 500 enterprise leaders in the US and UK found only 21% are seeing a significant, positive return on their AI investments. The cause is not the technology, it is the near-total absence of people strategy. Organisations with mature, company-wide AI literacy programmes are nearly twice as likely to report significant ROI. The 79% are not just missing returns; many are actively damaging their businesses through inaccurate decisions, lost competitive ground, and uncontrolled shadow AI use.
What does the data actually say?
The numbers are blunt. DataCamp surveyed over 500 enterprise leaders across the US and the UK, not garage startups, major players, and found just 21% are seeing a significant, positive return on their AI investments. Four out of five businesses are getting nothing meaningful back.
The consequences are already showing up on balance sheets:
- 32% of leaders cited inaccurate decision-making as a direct result of poorly implemented AI
- 27% cited loss of competitive advantage
The very tool that is supposed to make businesses smarter is, for most, doing the opposite. It is not just failing to deliver, it is actively producing bad outcomes and handing advantage to competitors who got the approach right.
Why is the UK AI adoption rate misleading?
Adoption is accelerating fast. 54% of British firms now use AI in some form, up from 35% just a year ago. That headline looks like progress. It is not.
Less than a third of that 54% are seeing any positive return on their investment. More than half the businesses in the country have jumped on the bandwagon, and only a tiny fraction are actually getting anywhere. The rest are along for a very expensive ride.
Worse still, less than half of these companies even have a clear strategy for what they want AI to achieve. They have bought the tool, ticked the box, and never asked the most fundamental question: why? What problem are we solving? What outcome are we driving?
Adoption is a mile wide. Understanding is an inch deep.
This is a classic case of hype over substance. Treating AI like a magic wand, expecting to wave it and watch problems disappear, is not a strategy. It is a gamble with company money on a technology that no one in the organisation truly understands.
What separates the 21% from the 79%?
The DataCamp data is unequivocal: organisations with mature, company-wide data and AI literacy programmes are nearly twice as likely to report a significant positive ROI.
Twice as likely. That is not a small correlation. It is the clearest signal in the entire dataset about where the real value lies.
The companies winning with AI are not the ones with the biggest tech budgets. They are the ones with the smartest teams. AI is not a passive technology, it requires human guidance, human intuition, and human expertise to be effective. Your team needs to know how to ask the right questions, interpret results accurately, and, critically, recognise when the AI is wrong.
Without that literacy, you are letting a black box make decisions for your business with no way of knowing whether those decisions are brilliant or catastrophic.
Is investing in AI upskilling actually worth it?
It is not a cost. It is the single best investment you can make in your AI strategy.
A tool can be copied by a competitor overnight. A capability, deep-seated expertise embedded across your team, is a competitive advantage that is genuinely hard to replicate. It is the difference between buying a tool and building a muscle. The 21% understood from day one that the return on AI is directly proportional to the investment made in the people who use it. They are not lucky. They are smart.
Investing in upskilling creates a culture of curiosity and critical thinking where your team is not just using AI, but actively partnering with it to drive real, measurable results.
See where AI fits in your business. Free.
A 45-minute audit. We map the highest-value automations and what they're worth in time and money. No pitch, no pressure.
What does good look like? The Singapore model
Singapore is not talking about this, it is doing it. The government launched the Digital Leaders Accelerator Bootcamp (DLAB). Note the name carefully: not a "Digital Tools Bootcamp." A Digital Leaders Bootcamp. The focus is people, not software.
DLAB runs intensive programmes in partnership with global consulting firms, taking business leaders through practical AI implementation that delivers tangible value in under three months. It is not theory. It is building capability and muscle memory directly into the leadership layer of businesses.
The Singaporean government backs this with a 400% tax deduction on AI expenses, but it is tied to a clear framework for responsible implementation, including the Monetary Authority of Singapore's AI Risk Management Toolkit. It is not a free-for-all. It is spending smart. It is creating an ecosystem where AI adoption is driven by strategy, not hype. The UK and Australia could learn a great deal from it.
What happens when you get it wrong? Australia's warning
Australia is the cautionary tale for what a tool-first, people-last approach produces.
- 90% of Australian security teams are feeling pressure to loosen their security controls to accommodate the business demand for new AI tools
- 67% of Australian workers are already using unapproved "shadow AI" tools, two-thirds of a workforce going behind the organisation's back, plugging company data into unknown AI platforms with zero oversight, zero security, and zero strategy
This is the inevitable consequence of a top-down, tool-focused approach. Businesses create demand for AI, tell everyone it is the future, but provide neither the right tools nor the right training. So employees find their own. They are trying to do their jobs and be more productive, but in doing so, they expose the business to data leaks, compliance breaches, and reputational damage that could be severe.
You can't just throw technology at a problem. You have to bring your people along, or they'll find their own way, and you won't like where it leads.
Australia is a case study in what happens when you focus on the what, the AI tool, and completely neglect the how, the strategy, the training, the culture.
Are you actually in the 21%?
Be honest. You may already have a ChatGPT subscription for your team or have signed a vendor contract. You are officially in the 54% of UK businesses using AI. But are you in the 21%?
Ask yourself:
- Are you tracking the return on that investment in measurable, pound-and-pence terms?
- Do you have a written plan to build AI literacy across your team?
- Are you confident your employees are not already using shadow AI tools you know nothing about?
Hope is not a strategy. If you are not actively and deliberately investing in your people's ability to use these tools, you are leaving significant value on the table, falling behind competitors who are making that investment, and exposing the business to the same security risks currently playing out in Australia.
What to do this week
- Audit every AI tool you are paying for. List each subscription or licence. For each one, write down the specific, measurable outcome it was purchased to deliver. If you cannot write it down, you do not have a strategy, you have a purchase.
- Run a shadow AI check. Ask your team honestly, and without judgement, what AI tools they are currently using outside of official systems. The answers will tell you exactly where the gaps are.
- Map your AI literacy baseline. Can your team critically evaluate AI outputs? Do they know when to challenge a result? If you cannot answer yes with confidence, that is your first investment, not your next software licence.
- Define one clear AI outcome for the next 90 days. Not "use AI more." One specific, measurable business outcome. Accountability starts with a number.
- Build capability before you buy more tools. Before signing any new AI contract, ask: do we have the people skills to extract real value from this? If the answer is no, invest there first.
Where to from here
Book a free 60-minute AI audit, we'll explore exactly what workflows are worth augmenting with AI.
Live with passion & AI,
Brett
Host a podcast? Have Brett on as a guest.
Straight talk on implementing AI in real SMEs, no jargon, plenty of receipts from the businesses we run.
Frequently asked questions
What percentage of businesses are getting a return on AI investment?
+
According to a DataCamp survey of over 500 enterprise leaders across the US and the UK, only 21% are seeing a significant, positive return on their AI investments. That means four out of five businesses are getting little to nothing back.
Why is AI ROI so low for most businesses?
+
The primary cause is a lack of investment in people rather than technology. The DataCamp data shows organisations with mature, company-wide AI literacy programmes are nearly twice as likely to report significant positive ROI. Most businesses focus on acquiring tools while ignoring the human layer required to extract value from them.
How many UK businesses are currently using AI?
+
54% of British firms are now using AI in some form, up from 35% just a year ago. However, less than a third of that 54% are seeing any positive return on their investment, and less than half even have a clear strategy for what they want AI to achieve.
What are the consequences of poorly implemented AI?
+
DataCamp's survey found 32% of enterprise leaders cited inaccurate decision-making as a direct result of poorly implemented AI, and 27% cited loss of competitive advantage. The technology designed to make businesses smarter is, for most, producing the opposite outcome.
What is shadow AI and why is it a business risk?
+
Shadow AI refers to employees using unapproved AI tools without organisational oversight. In Australia, 67% of workers are already doing this, exposing businesses to data leaks, compliance breaches, and reputational damage, all without any strategy or security controls in place.
What is Singapore doing differently with AI adoption?
+
Singapore launched the Digital Leaders Accelerator Bootcamp (DLAB), running intensive programmes with global consulting firms to help business leaders implement practical AI projects that deliver tangible value in under three months. The government also offers a 400% tax deduction on AI expenses, tied to a responsible implementation framework including the Monetary Authority of Singapore's AI Risk Management Toolkit.
What does an effective AI literacy programme look like?
+
An effective programme builds critical AI skills across the entire organisation, not just the tech team, so employees can ask the right questions, interpret outputs accurately, and identify when the AI is wrong. It shifts AI from a purchased tool into an embedded organisational capability that competitors cannot easily replicate.
Brett is a four-time founder (Darra Tyres, Gladfish, EzyTrac, Anaboo) and the operator behind AIOS, Anaboo's AI Operating System. He writes from inside the build, installing AI in his own businesses first and reporting back what actually moves the numbers. Based between Singapore, the UK and Australia.
