ASIA
08 September 2026
Does Artificial Intelligence boost productivity?
The economic impact of Artificial Intelligence may not be as revolutionary as many imagine.
Artificial Intelligence (AI) with its rapidly expanding capabilities is emerging as a new General Purpose Technology (GPT), comparable to earlier digital technologies such as the internet and personal computers, or previous breakthrough innovations like the steam engine and electricity.
Generally, past inventions eventually led to periods of accelerated productivity growth. Can AI play a similar role now and revive productivity growth in the US and other advanced economies?
AI will contribute to annual Total Factor Productivity (TFP) growth in the US by around 0.25-0.6 percentage points, according to research by the OECD. These figures may seem small. But US annual TFP growth was around 1 percent (and even lower for the OECD countries overall) over the past two decades.
Thus Generative AI will likely be an important source of aggregate productivity growth over the next 10 years, even though current AI technology alone is unlikely to bring productivity growth back to the
levels seen in the 1960s.
The productivity boost from AI will also likely be lower than the latest technology driven boom linked to information and communication technologies (ICT) which has been estimated to have contributed up to 1-1.5 percentage points to annual US TFP growth during the 1995-2004 period. But the boost from AI could be comparable to that of the ICT boom if AI capabilities become applicable in a wider range of activities, notably thanks to further integration with robotics technologies.
The OECD offers some lessons about the productivity boost from AI.
First, high and widespread adoption across many sectors of the economy, along with expanded AI capabilities – notably via the development of complementary digital tools or further integration with robotics technologies – is a key driver of achieving higher aggregate TFP gains.
Second, a stronger AI concentration and larger disparity across sectors, as in the past, would limit the productivity gains from AI.
Third, if physical or manual tasks are also exposed to AI, notably thanks to further integration with robotics technologies, the productivity benefits will also be more widespread across sectors.
In sum, the macro-level productivity gains from AI depend on several conditions and should not be taken for granted.
As also stressed by the OECD Principles on Artificial Intelligence, governments have a key role to play in making the most of AI in the following areas.
1. AI diffusion and adoption: Governments can support the capabilities of firms to adopt AI in a wide range of sectors through educational and skills development policies and via improving access to digital technologies, including through liberalised digital trade.
2. Demand for AI powered goods and services: Governments, in partnership with the tech sector, should foster trustworthiness, which is key to ensure demand for AI powered goods and services will meet supply and thus enable broad-based macroeconomic productivity gains.
3. Reallocation of factors: Governments should support workers in transitioning between jobs and across sectors through re-training and other labour market policies.
Generally, past inventions eventually led to periods of accelerated productivity growth. Can AI play a similar role now and revive productivity growth in the US and other advanced economies?
AI will contribute to annual Total Factor Productivity (TFP) growth in the US by around 0.25-0.6 percentage points, according to research by the OECD. These figures may seem small. But US annual TFP growth was around 1 percent (and even lower for the OECD countries overall) over the past two decades.
Thus Generative AI will likely be an important source of aggregate productivity growth over the next 10 years, even though current AI technology alone is unlikely to bring productivity growth back to the
levels seen in the 1960s.
The productivity boost from AI will also likely be lower than the latest technology driven boom linked to information and communication technologies (ICT) which has been estimated to have contributed up to 1-1.5 percentage points to annual US TFP growth during the 1995-2004 period. But the boost from AI could be comparable to that of the ICT boom if AI capabilities become applicable in a wider range of activities, notably thanks to further integration with robotics technologies.
The OECD offers some lessons about the productivity boost from AI.
First, high and widespread adoption across many sectors of the economy, along with expanded AI capabilities – notably via the development of complementary digital tools or further integration with robotics technologies – is a key driver of achieving higher aggregate TFP gains.
Second, a stronger AI concentration and larger disparity across sectors, as in the past, would limit the productivity gains from AI.
Third, if physical or manual tasks are also exposed to AI, notably thanks to further integration with robotics technologies, the productivity benefits will also be more widespread across sectors.
In sum, the macro-level productivity gains from AI depend on several conditions and should not be taken for granted.
As also stressed by the OECD Principles on Artificial Intelligence, governments have a key role to play in making the most of AI in the following areas.
1. AI diffusion and adoption: Governments can support the capabilities of firms to adopt AI in a wide range of sectors through educational and skills development policies and via improving access to digital technologies, including through liberalised digital trade.
2. Demand for AI powered goods and services: Governments, in partnership with the tech sector, should foster trustworthiness, which is key to ensure demand for AI powered goods and services will meet supply and thus enable broad-based macroeconomic productivity gains.
3. Reallocation of factors: Governments should support workers in transitioning between jobs and across sectors through re-training and other labour market policies.