Artificial intelligence (AI) could generate $2.52 trillion in annual economic value globally, but higher deployment costs could significantly reduce those gains, according to a new report by consulting firm McKinsey.
The report, titled The AI Value Curve: The Economics of Applying AI to Existing Work, examines how AI can automate and improve existing work processes. It highlights a gap between the rapid adoption of AI and the economic benefits companies are actually capturing.
According to McKinsey, many organisations are prioritising AI implementation without adequately assessing whether the technology delivers sufficient returns on investment.
Rising AI deployment costs could erase $850 billion
McKinsey estimated that a 30 per cent increase in the effective cost of AI deployment could reduce its projected annual global economic value by as much as $850 billion.
Under this scenario, the estimated value would fall from $2.52 trillion to $1.67 trillion annually.
In the United States, annual economic value generated by AI could decline from $960 billion to $640 billion if deployment costs rise by a similar margin.
McKinsey clarified that these figures represent a modelled scenario rather than a prediction of future outcomes. The estimates highlight how implementation expenses could affect the financial returns businesses derive from AI adoption.
AI infrastructure investment could reach $7 trillion by 2030
The findings come as businesses increase spending on AI infrastructure, software platforms and automation technologies.
According to the report, global investment in AI infrastructure could reach approximately $7 trillion by 2030. Such spending would require annual profit pools of around $760 billion to justify the investments.
The projections underscore the importance of balancing infrastructure expenditure with measurable productivity gains and commercial returns.
For companies adopting AI, the economic benefits will depend not only on deploying the technology but also on controlling costs and integrating it effectively into existing workflows.
Financial services leads potential AI gains
Financial services emerged as the industry with the highest estimated annual economic gains from AI adoption, at $530 billion.
The education sector ranked second, with potential annual benefits of $410 billion. Technology, media and telecommunications could generate $360 billion, while healthcare could realise an estimated $310 billion in annual value.
The estimates point to potential applications across sectors, including process automation, productivity improvements and support for professional decision-making.
AI could augment highly skilled professionals
The report also challenges the view that AI’s main economic role is to replace lower-wage workers.
Instead, it identifies opportunities to support highly skilled professionals in areas such as law, healthcare, management and computer-related occupations.
In these fields, AI could enhance productivity and decision-making by assisting professionals with existing tasks.
The findings suggest that companies may need to assess AI investments based on the value created across entire work processes, rather than focusing solely on automation or workforce replacement. The report’s projections also indicate that controlling deployment costs will be critical to converting AI adoption into measurable economic returns.
