Predicted Surge in Spending on Generative AI, Primarily for Hardware

The global expenditure on IT products and services related to generative artificial intelligence (AI) is projected to reach $644 billion by 2025, according to analysts at Gartner. This figure represents an increase of over 75% compared to the previous year, indicating a significant shift in the IT spending landscape.

Gartner forecasts that generative AI will have a profound impact on all aspects of IT expenditure, becoming an integral part of business operations and consumer products. However, rather than a widespread uptake of AI service subscriptions, the bulk of the spending--approximately 80%--is expected to be directed towards hardware that has been enhanced with AI capabilities.

Specifically, spending on AI-enabled servers is anticipated to rise from $135 billion in the prior year to $180 billion, marking a 33% increase. Furthermore, expenditures on other AI-capable devices such as smartphones and personal computers are projected to nearly double, reaching close to $399 billion.

In contrast, the anticipated spending on software and AI services appears modest, with estimates of around $37 billion and $27 billion, respectively. Yet, both areas are expected to see explosive growth, with service spending likely increasing by 162% and software by nearly 94%.

According to Gartner analyst John-David Lovelock, the market growth is heavily influenced by the rising prevalence of AI-capable devices. By 2028, these products are expected to dominate the consumer device market. Lovelock notes that consumers are not actively seeking these AI functionalities, but manufacturers are making them standard features, effectively compelling consumers to purchase them.

Despite the push for AI integration, Lovelock suggests that companies are facing challenges regarding the reliability of AI technology. High error rates and dissatisfaction with the outcomes of proof-of-concept projects have led to diminished expectations. Nonetheless, companies continue to invest billions annually to enhance the size, performance, and reliability of generative AI models.

Moreover, Lovelock anticipates that businesses will reassess their ambitious internal AI projects, with some potentially being shelved. Instead, Chief Information Officers (CIOs) are likely to shift toward commercially available standard solutions from various providers, which are perceived as offering more predictable implementation and greater business value.