Businesses are investing in AI faster than they can make it pay, new TPI research finds

AI is making individual tasks much faster, but the productivity gains are not automatically translated to the bottom line. TPI research for Bytes Technology Group says closing the UK’s “AI ambition gap” could add 0.5 to 0.7 percentage points a year to productivity.

• The UK is well placed: of six countries compared, it has the largest share of workers whose jobs are most likely to be helped rather than replaced by AI
• Faster tasks do not automatically mean more productive businesses: gains depend on redesigning whole workflows, rather than simply adding AI to individual activities
• Closing the gap requires investment beyond technology: management, skills and organisational change are essential to turn AI’s potential into sustained business results

New research from The Productivity Institute (TPI) finds that businesses are investing in artificial intelligence far faster than they can turn it into real productivity gains. The research was commissioned by Bytes Technology Group, a UK FTSE 250 technology company.

The report – The AI Ambition Gap – brings together the latest evidence from around the world on AI and productivity, alongside interviews with experts from business, academia and government. It finds that AI makes individual tasks dramatically quicker, but those gains shrink every time AI steps up one level from task to a whole job, from team to the organisation, and from organisation to the wider economy.

“AI is making individual tasks much faster, but the productivity gains are not automatically translated to the bottom line. Closing the UK’s “AI ambition gap” could add 0.5 to 0.7 percentage points a year to productivity.”

Where the gains leak away

Studies reviewed in the report show customer-support staff resolving 14% more customer queries an hour with an AI assistant, with the biggest gains for less experienced staff. Professionals using AI finished writing and analysis tasks 40% faster, and to a higher standard.

Across whole organisations, however, AI use is still shallow. Nearly nine in ten (89%) organisations use AI somewhere in the business. Most workers use it for no more than one to two hours per week. Only one in four has changed how work is done to make the most of it. Fewer than half (46%) of organisations have got beyond pilot projects, and only 6% of firms are getting 5% or more of their profits from AI.

Software development shows the problem clearly. Developers using AI coding tools produce more than three times as much code, but release only 30% more software. The gap is due to the need for people who still have to check, test and launch it.

Management, not technology

The researchers identify an “AI ambition gap”: the difference between what AI could do for a business and what it actually delivers. They conclude that closing this gap requires leaders to set a clear strategic direction, create incentives for change, and equip staff with the skills and confidence to try things safely. Above all, leaders must redesign whole workflows so that faster individual tasks translate into better results across the business.

Good management is the clearest dividing line. UK companies that were better managed in 2020 were more likely to be using AI by 2023, regardless of their size, age, sector or location. Differences in management practices, together with workforce and company characteristics, explain around 80% of the gap between the shares of US and European workers using generative AI at work.

The report also explains why AI can look as if it is not paying off. Companies spend heavily up front on retraining staff, reorganising work and sorting out their data. Those costs show up straight away, but the payback comes later, so productivity can dip before it rises. Economists call this the “J-curve”.

The prize for the UK

Productivity growth across advanced economies has roughly halved since the mid-2000s, from around 2% a year to about 1%. The researchers judge that AI could most plausibly add 0.5 to 0.7 percentage points a year to UK productivity growth, which they call “a significant step change”.

The UK is well placed to benefit. The report cites research comparing the UK with five other countries (the US, Brazil, Colombia, India and South Africa), showing it has the largest share of workers in jobs where AI has considerable scope to complement human expertise. This partly reflects its large financial and professional services sectors, although realising these benefits requires appropriate skills and organisational support.

The report says the quickest win for the UK and Europe is not building new AI tools but putting existing ones to work in their strongest industries, such as health, professional services and advanced manufacturing.

Sam Mudd, CEO of Bytes Technology Group, said, “AI takes extraordinary capital to build and almost nothing to try, so it is arriving in every organisation whether leaders have a plan or not. Across our customer base, companies in the same sector are making completely opposite bets, from licences for every employee to nothing at all. Neither guarantees results. This research shows that the organisations pulling ahead are the ones treating AI as a business change, with clear goals, safe space to experiment and the skills to scale what works.”

Professor Bart van Ark, Senior Fellow at TPI, said, “AI’s productivity potential becomes real when faster tasks lead to better performance across the business. The ambition gap lies between those individual efficiencies and sustained business results. Closing it requires firms to invest in management, skills and the redesign of workflows alongside the technology.”

Closing the gap

The report sets out seven enablers to realise the AI ambitions. For business leaders: set a clear goal for what AI should achieve, train people properly, connect technical feasibility with economic viability, and be ready to change how work is done. For government: clear rules tailored to each sector, better local support for businesses, and stronger links between AI research and the companies putting it to use.

Read the full report: The AI Ambition Gap: Aligning Capabilities for Productivity