The acceleration of mass adoption for generative artificial intelligence (GenAI) is expected to boost revenues from this technology almost 20-fold over the next three years.
Morgan Stanley research predicts that software and internet companies will experience a more than 20-fold increase in revenue from Generative AI over the next three years, from $45 billion in 2024 to about $1.1 trillion in 2028, achieving a positive return on investment already this year, with an estimated margin of 34%.
As Generative AI’s capabilities continue to expand, driving wider adoption, it is expected to fuel a new technology cycle and significantly increase both related profits and expenses. Thus, GenAI semiconductor spending is forecasted to reach $280 billion in 2028, while corporate spending on hardware, networking, and memory for GenAI is predicted to almost triple, up from $98 billion in 2024 to $276 billion in 2028.
Generative AI attracted mass attention in November 2022 with the launch of ChatGPT, followed by the release of similar business-focused products in late 2023. Initially, it was unclear when or if the technology would become profitable. However, as development advanced and adoption grew, its revenue potential became more evident.
For 2025, Morgan Stanley Research forecasts $153 billion in revenue, with enterprise software bringing in $59 billion and consumer platforms generating $94 billion.
For the first time, Generative AI is expected to deliver a positive return on investment (ROI) this year, with a projected 34% contribution margin, or $51 billion, compared to a -5% margin last year. By 2028, GenAI ROI is expected to rise further, reaching a 67% margin and $722 billion in returns.
Out of $1.1 trillion in revenue projected for GenAI in 2028, $400 billion will come from businesses investing in automation-driven productivity software. Consumer platforms, including e-commerce portals, search, social media, autonomous technology, and wearables, could contribute an additional $680 billion by streamlining consumer processes.
One of the main factors that will spur GenAI growth is the level of savings this technology brings. Today, most major businesses are partially substituting their call centre services by GenAI assistants, which significanlty reduces operational costs.
As for the bottlenecks in the GenAI adoption, one must mention hardware challenges. An AI model that is just 10% smarter might require five times more hardware, according to Joe Moore, Head of U.S. Semiconductors Research at Morgan Stanley. Therefore, despite the recent development of some Chinese AI models that require smaller amounts of hardware, major companies continue to increase their spending plans as they scale up GenAI capabilities.
Thus, GenAI-related semiconductor spending is forecasted to reach $280 billion in 2028, up from $115 billion in 2024. Spending on hardware, networking and memory for GenAI is likely to surge threefold in the same period, to $276 billion in 2028, from $98 billion in 2024.
Another recent report revealed that GenAI and cloud computing growth is expected to double the capacity of data centers located in the Asia-Pacific region by 2028. The indicator is forecasted to increase by an average of 20% per year and reach about 24,800 megawatts (MW) in three years, up from the 10,500 MW registered in 2024.
