Goldman says investment in AI could near $200bn by 2025

The potential of Generative AI to significantly enhance labor productivity and overall economic output is substantial, particularly as it gains widespread adoption across various industries

An analysis of global business trends indicates that investments in artificial intelligence (AI) technologies by corporations worldwide could approach a collective sum of $200 billion by 2025, as forecasted by economists at Goldman Sachs.

The potential of Generative AI to significantly enhance labor productivity and overall economic output is substantial, particularly as it gains widespread adoption across various industries. The economic study, conducted by Goldman Sachs economists Joseph Briggs and Devesh Kodnani, projects a substantial surge in private AI investment on a global scale over the next three years. Starting from an estimated $91.9 billion in 2022, the projection indicates a steep rise to exceed $158.4 billion in 2025, marking a 72% increase within this three-year timeframe.

This growth trajectory is underlined by the economists’ observation that AI investment has risen from a relatively modest point. Measured through the revenues of prominent players such as Microsoft Azure, Nvidia, Google Cloud, and Amazon Web Services, global AI investment saw an ascent from approximately $3.2 billion globally in 2013 to around $25.5 billion in 2017. It further reached approximately $48.2 billion in 2020 and surged to $93.5 billion in 2021. Following a temporary levelling in the subsequent year, a rebound is anticipated.

Goldman Sachs’ report predicts consistent upward momentum in AI investment on a global scale over the coming years, projecting figures to reach $110.2 billion in 2023 and then escalating by over $20 billion annually, culminating in the forecasted $158.4 billion by 2025.

In terms of geographical distribution, private AI investment in the United States accounted for more than half of the global figure in 2022, totalling $47.4 billion out of the $91.9 billion worldwide total, representing approximately 51.6% of the entire AI investment.

Looking ahead, Goldman Sachs’ forecasts propose a rise in U.S. AI investment to $56.8 billion in 2023, $68.1 billion in 2024, and subsequently reaching $81.7 billion in 2025. The U.S.’s proportion of the global private investment is anticipated to remain relatively constant in the upcoming years.

For comparison, China’s AI investment amounted to $20.6 billion in 2022 and is predicted to ascend to $35.6 billion in 2025.

The report’s authors discern that a significant portion of AI investment will likely be concentrated in four key business sectors: companies specialising in AI model development or training; entities providing infrastructure to support AI applications, such as data centres; software development firms enabling AI utilisation; and corporate end-users investing in software and cloud infrastructure services.

Briggs and Kodnani state, “While AI investment thus far has been focused on model development, a substantially larger hardware and software push will likely be required for generative AI to scale.”

Although AI investment is set to experience rapid escalation in the foreseeable future, it will still represent a minor proportion of both the U.S. and global economic output.

The report suggests that despite the rapid growth, the immediate impact on GDP is expected to be relatively moderate due to the current low share of AI-related investment in the U.S. and global GDP.

The research team at Goldman Sachs projects that AI investment in the U.S. could peak at over 2% of the country’s gross domestic product (GDP) before entering a declining phase.

Long-term projections suggest that AI investment might eventually reach 2.5% to 4% of GDP in the U.S., while in other leading AI nations, it could peak within the range of 1.5% to 2.5% of GDP.