AI is inescapable. Every article, it seems, mentions it. Just about every TV announcer brings it up. There is hardly a vendor in any industry that doesn’t claim to be implementing it. On the consumer side, most people have at least tried it out. So, is it all hype? Or is there real substance behind it?

According to Forrester Research, the answer is a bit of both. Forrester expects AI spending to slow in 2026 and predicts turmoil in the market, as well as continued vendor fragmentation.

“In 2025, AI agents were a hot topic in the data and analytics world, but the initial hype is expected to fade in 2026, giving way to a more practical focus on AI governance, literacy, and using agents for routine data tasks,” said Sudha Maheshwari, an analyst at Forrester Research. “The excitement around what’s possible with AI will be replaced by the reality of its everyday applications.”

Here are some of the top trends outlined by Forrester Research:

More Cautious Spending

In 2025, many organizations said they were going to spend big on AI. But less-than-promising results in some cases have resulted in a more cautious approach. As a result, many enterprises will delay 25% of planned AI spending until at least 2027, according to Forrester. The reason? AI value is failing to materialize as expected, with only 30% of organizations able to directly tie the value of their AI initiatives to increased profit, sales, or efficiency. CFOs, who lost some control over AI spending in 2025, have returned to the table and are being included in more AI deals. With finance teams more involved, spending will slow and many projects will be delayed. Further, a quarter of CIOs will be asked to bail out business-led AI failures.

Vendor Fragmentation

Hyperscalers, data platform vendors, and automation vendors can’t claim agentic AI dominance just yet. With so many vendors offering fragmented solutions, enterprises are being forced to compose “agent-lakes” to manage and orchestrate fractured AI agent deployments and enable complex multi-agent use cases across multiple solutions and platforms. Vendor consolidation will eventually happen, but not for a while.

AI Governance

A big problem with initial AI projects was a lack of control. According to Forrester, 60% of Fortune 100 companies will appoint a head of AI governance. Many organizations outside the Fortune 100 will do the same. This is being driven by the complexities of navigating regional and global legislation. The EU AI Act, for example, is expected to push many Fortune 100 companies to establish AI governance leadership roles. Meanwhile, some U.S. states are proposing their own laws.

“Every bubble inevitably bursts, and in 2026, AI will lose its sheen, trading its tiara for a hard hat,” said Maheshwari. “Companies will distribute their bets across agentic ecosystems and shift talent around as AI agents take over grunt work.”

What Does This Mean for Data Centers?

What do these trends mean for data centers? Gone may be the endless frantic meetings between data center personnel and management about how the organization must adopt AI immediately. More pragmatic voices are emerging that demand tangible value from AI investments. Data center managers may no longer be told to randomly redesign aisles or facilities to make way for AI applications and GPUs. Before money is spent, the return-on-investment question will need to be answered.

In addition, CFOs and CIOs are likely to return their attention to data center efficiency. Instead of spending heavily based on AI hype, they will expect to see value from investments already made. They will be asking data center managers about PUE and efficiency. They will want to know that the data center is addressing airflow fundamentals effectively. Before implementing AI broadly, they will want confidence that a solid data center foundation exists on which to build the AI enterprise.