AI to Take 20% of IT Budgets by 2027, Forrester Warns

What happened

Research firm Forrester expects artificial intelligence to account for 20% of corporate technology spending by 2027. The projection, reported by The Register and Computerworld, comes with a warning that more than half of IT executives are likely to miss their return-on-investment targets.

The survey findings point to a gap between expectations and results. Forty-five per cent of AI decision-makers say an investment must deliver a positive return within one year to count as a success, yet only 22% report a positive effect on their company's operating profit over the past year.

Forrester links part of the problem to organisations adding AI technology without changing their structures or work processes, so financial gains tend to appear within isolated use cases rather than across the business. The firm expects companies to rely more on open AI models, which are cheaper to run, and to reserve more expensive, advanced models for the most critical tasks.

Why it matters

AI is moving from experimental budgets into core technology spending, which raises the bar for proving value. The Forrester findings suggest that the technology is rarely the constraint. The constraint is whether an organisation redesigns the workflows around it. A tool deployed into an unchanged process can improve one task while leaving the end-to-end journey, and the cost to serve, largely untouched.

For leaders in experience and digital transformation, this shifts the question from "Which AI tools should we buy?" to "Which service journeys should we rebuild, and how will we know they work?" Model selection also becomes an operating decision, as cheaper open models and premium models carry different cost profiles and suit different levels of risk.

By the numbers

• 20% of corporate technology spending projected to go to AI by 2027, according to Forrester.
• More than half of IT executives expected to miss their return-on-investment targets.
• 45% of AI decision-makers require a positive return within one year to consider an investment successful.
• 22% have seen a positive impact on their company's operating profit over the past year.

The Renascence take

The one-year payback test is probably measuring the wrong thing. Most AI value sits in how a journey changes, not in how a single tool performs, so judging each investment on its own short-term return will tend to reward the wrong projects.

Customers do not adopt software; they adopt changed routines, and routines only shift when the process around them is redesigned. Judge AI pilots on journey outcomes such as resolution time, effort and repeat contact, not on tool-level savings. Reserve premium models for the moments where a poor answer damages trust, and let cheaper models handle the routine volume. If your AI budget is growing faster than your service metrics, the problem is design, not spend.