Artificial intelligence is starting to show up on corporate bottom lines, but a new study suggests that most companies are still struggling to move beyond small-scale experiments. According to research from management consultancy BearingPoint, nearly three-quarters of surveyed companies have seen some financial benefit from AI, yet only 13% are on track to deploy it widely across their operations.
The findings paint a picture of an AI boom that is real but uneven. While many firms are reaping cost savings, far fewer are using the technology to generate new revenue. And the barriers to broader adoption are less about the technology itself and more about the practical realities of regulation and aging IT infrastructure.
What the study found
BearingPoint surveyed companies across multiple industries and countries. The headline number: 13% of firms are considered "on track" to scale AI effectively. That means they have moved beyond pilots and are integrating AI into core business processes.
The gap between seeing some benefit and scaling it is stark. While nearly 75% of companies reported some financial upside from AI, fewer than a third have rolled it out broadly. The study also found that 24% of companies reported AI-driven cost savings of at least 10%, but only 4% saw revenue growth of that magnitude. That suggests most firms are using AI to automate existing tasks rather than to create new products or services.
"The early payoff from AI is showing up more in payroll costs than in sales," the study's authors noted.
Why scaling is so hard
The biggest obstacles are not theoretical. About 40% of respondents pointed to legal and regulatory constraints as a major hurdle, while 34% cited the difficulty of integrating new AI tools with existing legacy systems. These are practical, day-to-day challenges that can stall even the most promising projects.
Adoption also varies significantly by country. In China, 20% of companies reported comprehensive AI implementation, and in the US the figure was 18%. Germany lagged behind at just 8%. This geographic split reflects differences in regulatory environments, corporate culture, and the state of existing IT infrastructure.
Despite these challenges, there are signs of progress. The share of companies saying AI is "deeply integrated" into their operations rose to 11% in 2026 from 7% in 2025. That's still a small minority, but the trend is upward.
What it means for investors
For everyday investors, the study offers a window into how companies may think about AI in the near term. The pattern of cost savings outpacing revenue growth is telling. When managers see AI as a way to cut costs rather than to drive sales, they tend to treat it as a margin lever. That can translate into slower hiring, frozen backfills, and more pressure on back-office roles.
Nearly two-thirds of companies said they have excess staffing of at least 10%, according to the study. That, combined with the fact that far more firms reported significant cost savings than revenue growth, suggests that AI is being used to trim headcount rather than to expand into new markets.
For investors, this means that even "successful" AI programs may not lead to top-line growth. Instead, they might show up as improved profit margins, which can be positive for earnings but may also signal a more cautious approach to hiring and investment.
The study also highlights the importance of regulatory and IT infrastructure. Companies that can navigate these hurdles may be better positioned to scale AI and reap the benefits. Those that can't may remain stuck in pilot mode, missing out on the full potential of the technology.
As AI continues to evolve, investors will be watching to see which companies can move from experimentation to broad deployment. The ones that do may see a more significant impact on their financial performance. For now, the study suggests that the AI revolution is still in its early stages, with most companies yet to unlock its full value.
