Firms pour billions into AI but struggle to scale value

Corporate investment in artificial intelligence rose sharply last year, but most organisations still struggle to turn that spending into measurable results at scale, according to new research from Economist Impact.

The study found that 88% of senior executives view AI as a competitive advantage, yet only 4% said their organisations had achieved repeatable, scalable business value from it. It put global corporate AI spending at USD $252 billion in 2024.

The findings are based on a survey of 639 senior executives in London, New York, Tokyo, Singapore and Sydney, spanning a range of sectors and roles including technology, HR and general management.

Productivity focus

Short-term productivity gains dominated the business case for AI. Some 73% of executives cited employee productivity as the primary driver of AI investment, and 40% described AI as a tool to augment human roles and free workers for higher-value tasks.

That emphasis also shaped how firms measured performance. The research found that 79% assessed AI success using productivity metrics, while far fewer tracked longer-term measures such as employee engagement, skills development or retention.

Approaches varied by location. In Tokyo, 88% of executives said employee productivity was the top return-on-investment metric for AI investments-the highest share among the cities surveyed.

Charles Ross, Economist Impact's Asia head of policy and insights, said the pace of AI deployment is outstripping organisational readiness.

"Our research indicates that many businesses are still laying the tracks while the AI train is in full motion," Ross said.

"Driven by the desire for quick, tangible wins, organisations often equate productivity gains with ROI. Without clear strategy, skills and governance, they risk missing the sustainable competitive advantage AI can deliver," he added.

Uneven maturity

More than two-thirds of executives said their organisations had moved beyond AI experimentation. Even so, only a small minority reported achieving repeatable value across the organisation.

New York recorded the highest share of firms reporting scalable AI impact, at 6%, followed by Tokyo and London at 5%. Singapore stood at 4%, while Sydney recorded 0%.

The research also highlighted gaps in accountability and governance. While all organisations surveyed had discussed or planned frameworks for responsible AI use, only 8% said they had a comprehensive, actively enforced AI governance framework. Among small firms, that figure fell to 2%.

Enforced governance also varied by city. Tokyo reported the highest rate at 11%, followed by New York at 10% and London at 8%. Singapore and Sydney reported 5% and 4%, respectively.

Governance risks

Ross said internal practices create some of the biggest risks as AI use expands.

"Effective AI adoption depends as much on human vigilance as on technology," he said.

"The greatest risks often stem from internal missteps-poor data handling, weak oversight or unguarded use of sensitive information. Strong governance and skills development are essential to prevent AI from introducing new vulnerabilities to businesses and employees," he added.

Those vulnerabilities sit alongside widespread concerns about workforce readiness. The research found large gaps between the importance executives place on certain skills and their confidence in employees' proficiency.

Cybersecurity showed the biggest disparity: 96% of executives rated it as essential for AI deployment, but only 20% believed their teams were proficient. The study also found a 68-point gap in data privacy and a 71-point gap in bias detection.

Training gaps

Many organisations recognise the need to build AI skills but are not funding or structuring training widely. While 88% of executives viewed AI skills development as a competitive advantage, only 38% said they had dedicated and sufficient budgets for AI-related training.

Almost all respondents (99%) reported having some approach to developing AI skills. The most common methods were mentorship (54%) and self-directed online learning (52%). Structured internal programmes were less common (16%), while external partnerships stood at 21%.

Even where training exists, it often reaches only a small share of staff. Nearly half of executives said AI training covered less than 10% of their workforce.

London stood out for its emphasis on more formal talent pipelines. Some 53% of executives there cited university partnerships as a top strategy, compared with a global average of 43%. The survey also found that 41% of London firms were expanding successful AI initiatives across multiple business units.

Soft skills

The research also pointed to soft skills as a constraint on AI adoption and organisational change. Executives ranked critical thinking and creativity as highly important (both at 95%), but only around one-third believed employees currently excel in these areas.

Responsibility for skills development also appeared diffuse. Nearly half of executives said managers held minimal responsibility for developing AI skills, while 8% reported no responsibility at all. One in three leaders cited resistance to change among employees and middle management as a key barrier to aligning talent strategy with AI goals.

Keisuke Koyama, senior general manager of the Corporate Marketing Division at Kyocera Document Solutions, said the findings reflect a broader debate about how businesses assess AI investment.

"Organisations that prioritise short-term productivity over long-term skills development risk missing AI's true potential. This research highlights the non-technical factors-skills, governance, leadership-that determine whether AI ambition translates into sustainable business outcomes. Bridging these capability gaps is essential to responsible, people-centered digital transformation that delivers real business value," Koyama said.