A new laptop, a standard software update, or a server upgrade. There’s a good chance they will all incorporate AI soon, even if you’re not actively looking for it. According to a new report from Gartner, global spending on generative AI (gen AI) is expected to rise to $644 billion by 2025. That’s an increase of 76.4% compared to 2024. Companies are investing heavily. But many projects fail. What does that mean for you?
Hardware consumes the largest part of the AI budget.
Gartner's analysis shows that 80% of all AI spending goes to hardware. This is primarily because manufacturers are increasingly integrating standard AI functionality into their devices. Think of servers and devices that are 'AI ready.' In numbers:
Devices: $398.3 billion (99.5% growth)
Servers: $180.6 billion (33.1% growth)
Software: $37.2 billion (93.9% growth)
Services: $27.8 billion (162.6% growth)
It is notable that this hardware focus is not temporary. Hardware is expected to become even more dominant in the coming years. Much AI functionality will be integrated into software as a standard, without being separately visible in the budget.
The impact on budgets and strategies
Companies that believed they would need to invest less in hardware in the long run must reconsider their IT strategy. Generative AI functionalities are increasingly being delivered as part of existing software packages. This requires different planning and financial models. Additionally, it is becoming more challenging to set up standalone AI projects. AI is becoming more of an integral part of systems rather than a separate initiative.
Why AI Projects Keep Stalling
Although many organizations invest in AI, not every project succeeds. Gartner identifies three main causes of failure:
Poor or insufficient data
Lack of willingness to change among employees
Insufficient return on investment (ROI)
Experienced organizations achieve success more often. However, companies that are less familiar with AI more frequently get stuck in the so-called PoC phase (Proof of Concept).
The shift towards ready-made solutions
Instead of developing AI solutions themselves, an increasing number of companies are opting for commercial software with built-in AI features. These are quicker to implement, more predictable in terms of results, and often provide clear added value. CIOs and IT leaders would do well to approach AI as an integrated part of existing systems, rather than something separate.
This calls for an evolutionary approach.
The expected spending growth towards $644 billion is impressive. However, success is not determined by how much you spend, but by how well your organization is prepared. Companies that integrate smartly, involve their people, and critically examine data and ROI have an advantage. AI is on the way. The question is: will you use it as a hype, or will you truly implement it strategically?