AI Could Drive 3.8% Productivity Growth: IMF

Artificial intelligence could increase aggregate labour productivity by as much as 3.8 per cent over the long term, according to a new International Monetary Fund working paper that analysed patent and employment data across developed economies.
The study found that the rapid growth of AI-related innovation has already generated measurable productivity gains. Researchers estimated that AI patent activity between 2000 and 2017 increased output per worker by between 0.8 per cent and 1.2 per cent.
The findings appeared on Friday in an IMF working paper titled “Artificial Intelligence and Aggregate Labor Productivity: Evidence from Patent Data.”
Researchers examined the broader impact of AI technology on labour productivity using patent data from countries belonging to the Organisation for Economic Co-operation and Development.
The study covered the period from 2000 to 2017. During this period, AI-related patent issuance accelerated significantly and more than tripled by 2017.
OECD countries accounted for about 89 per cent of all AI-related patents issued worldwide during the period, according to the paper.
Using a production-function approach, the researchers assessed whether the growing number of AI patent applications translated into higher output per worker.
Their estimates indicate that AI innovation could generate considerably larger productivity gains in the future as technological development accelerates and businesses adopt AI across more industries.
However, the potential increase of up to 3.8 per cent represents a long-term estimate. The researchers noted that achieving these gains would depend on continued technological development and broader adoption of AI across different sectors.
AI is also unlikely to generate the same productivity gains across all countries and employment categories.
The study found that countries with a larger share of workers in professional and managerial roles recorded greater productivity gains from AI technology.
This finding indicates that AI may generate stronger economic returns when it complements skilled workers rather than simply replacing existing tasks.
However, businesses and workers may need time to realise these benefits. Employees must learn to use new AI systems and integrate them effectively into their daily responsibilities, the paper noted.
The speed at which workers and businesses adapt to AI could therefore determine how quickly the technology translates into higher output.
Labour-market flexibility could also affect the scale of productivity gains. Greater worker mobility could help employees move into occupations and industries where businesses can use AI more effectively.
The researchers said their conclusions remained robust even after accounting for technological spillovers between countries. Such spillovers occur when businesses or economies adopt and adapt AI advances developed elsewhere.
The findings highlight the link between technological innovation, workforce composition and economic policy. Creating more AI patents alone may not deliver the largest economic gains unless countries also develop a workforce capable of using the technology and maintain labour markets that support occupational mobility.
For India, these issues could become increasingly important as businesses expand their use of AI. The country has placed growing policy emphasis on artificial intelligence, digital public infrastructure and technology-led economic growth.
Its large information technology sector and skilled workforce could make employee training, productivity and workforce mobility key considerations in the wider adoption of AI.
The global discussion around AI has also expanded beyond technological capabilities to its potential effects on jobs, wages and economic growth. Governments and businesses are examining how investments in skills, research and digital infrastructure can help economies capture productivity gains while managing potential disruption across the workforce.









