Only 13% of organisations scale AI as planned, BearingPoint finds

Artificial intelligence is delivering measurable financial benefits for businesses, but only 13% of organisations have managed to scale their AI initiatives fully in line with their original business case, according to new research from BearingPoint.

The management and technology consultancy’s Scaling AI for measurable impact study found that nearly three-quarters of organisations that have implemented AI are already seeing a measurable impact on revenue or costs, with around four in ten reporting both revenue growth and cost reductions.

However, almost three-quarters have either adjusted the original scope of their AI programmes or achieved less scale than anticipated, highlighting a growing gap between demonstrating value through individual use cases and deploying AI successfully across an organisation.

The research surveyed 1,050 C-suite executives and senior leaders across 13 countries in Europe, the US and China.

“Organisations have shown that AI can create real business value, but proving value and scaling value are two very different things.

“We found that the organisations pulling ahead are not simply investing more. They are connecting AI to financial accountability, trusted data, governance, architecture, and workforce decisions from the start. Management discipline is what turns isolated success into organisational impact.”

AI maturity continues to increase

BearingPoint categorised respondents into four stages of AI maturity. Some 15% were classed as Explorers, which are still assessing potential AI applications, while 20% were Experimenters with active projects or pilots.

More than half – 54% – were Implementers that have deployed AI and are generating value, while 11% were classed as Leaders, where AI is deeply integrated into operations and its impact is measured against a defined transformation roadmap.

The proportion of organisations with AI deeply integrated across their operations has increased from 7% in 2025 to 11% this year.

Financial benefits are also expected to become more significant. Among 685 organisations that have implemented AI, just 4% currently report revenue or service-delivery gains of at least 10%. By 2030, 22% expect to achieve gains at this level.

Cost savings are already more widespread, with 24% reporting reductions of at least 10%. This rises to 35% expecting cost reductions of 10% or more by 2030.

Despite uncertainty over returns, nine in ten organisations said they would continue investing in AI even if expected ROI remained limited.

Scaling becomes an organisational challenge

The research identified complex regulatory frameworks as the leading barrier to scaling AI, followed by integrating the technology with legacy systems and processes.

More than half of executives – 54% – identified high-quality, trusted data as critical to successful scaling, alongside connected data, governance and ownership, and accessibility.

However, fewer than one-third of organisations formally assess scalability before beginning an AI initiative.

The difference becomes particularly pronounced between organisations at different maturity levels. Almost half of those classified as Leaders scale AI initiatives fully as planned, compared with just 6% of Implementers.

Seventy per cent of Leaders also connect most AI projects to measurable financial KPIs, compared with 34% of Implementers.

AI creates workforce capacity challenge

The study also highlights the impact AI is beginning to have on workforce planning.

Some 62% of organisations estimate that AI has already created workforce overcapacity of at least 10%, rising to 94% expecting overcapacity at this level by 2030.

At the same time, businesses increasingly require new capabilities in areas including AI governance, agent orchestration, data science and the design of workflows combining people and AI.

Gigant said: “AI adoption releases excess capacity, but management decisions determine whether that capacity becomes growth, or simply unused effort.

“Organisations need to redesign roles, redeploy resources, build new skills, and align workforce planning with their AI roadmaps. Otherwise, productivity can improve without producing the financial impact executives expect.”

Agentic AI is another area where ambition is currently running ahead of organisational readiness. More than three-quarters of respondents are still learning about agentic enterprise architecture, identifying it as a priority or exploring pilots. Only 13% have a defined strategy with active initiatives, while 10% are scaling agentic AI across their organisation.

BearingPoint concluded that businesses need to move from treating AI as a collection of individual use cases towards managing it as an enterprise-wide value portfolio, connecting investment decisions with financial outcomes, scalability, workforce planning, architecture and governance.

The research was conducted through online interviews in August 2026.