Global Spending on Generative AI Expected to Reach $14.2 Billion by 2025
Projected Growth in Generative AI Spending
New Delhi, July 10: A recent report from Gartner indicates that global expenditure by end-users on generative AI (GenAI) models is anticipated to hit $14.2 billion by the year 2025.
This year, spending on specialized GenAI models, which encompass domain-specific language models (DSLMs), is projected to reach $1.1 billion.
These specialized models are tailored or fine-tuned using data specific to particular industries or business processes.
According to Gartner's forecasts, by 2027, over 50% of GenAI models utilized by businesses will be domain-specific, a significant increase from just 1% in 2024.
“Foundation GenAI models, including large language models (LLMs), are developed using extensive datasets and are versatile for various tasks. They represent the initial models supporting GenAI and will continue to account for the largest share of organizational spending in the upcoming years,” stated Arunasree Cheparthi, Senior Principal Research Analyst at Gartner.
Organizations are increasingly adopting more domain-specific GenAI models due to their enhanced performance, cost-effectiveness, reliability, and relevance for targeted enterprise applications, she added.
A previous Gartner report projected that global spending on generative AI could reach $644 billion by 2025, reflecting a remarkable increase of 76.4% from 2024. This surge in GenAI expenditure will primarily be driven by the incorporation of AI capabilities into hardware, including servers, smartphones, and PCs, with 80% of the spending allocated to hardware.
Significant growth in GenAI spending is expected across all major markets and submarkets in 2025. The report suggests that GenAI will profoundly influence all facets of IT spending, indicating a future where AI technologies become increasingly essential to both business operations and consumer products.
Providers of foundational models are investing billions each year to improve the size, performance, and reliability of GenAI models. This trend is expected to continue through 2025 and 2026.