AI Could Raise Enterprise IT Costs by 75%, Bain Warns

What happened

Bain & Company has forecast that artificial intelligence could drive enterprise IT budgets up by as much as 75% within less than ten years, even where organisations pursue AI investment carefully. The consultancy points to rising infrastructure demands, heightened security requirements and the cost of specialised talent as the primary drivers behind the projected increase, according to reporting by CIO Dive.

The analysis suggests that the cost pressure is not confined to compute and cloud spend alone. Securing AI systems and hiring or retaining the skilled staff needed to build, govern and maintain them are expected to add substantially to the total bill, even for organisations that approach AI deployment with discipline rather than unchecked expansion.

Why it matters

For technology and transformation leaders, the finding complicates the simple narrative that AI automatically pays for itself through efficiency gains. Even disciplined, well-governed AI programmes appear likely to raise the underlying cost base of IT, which has direct implications for how CIOs build business cases, set budgets and sequence adoption across the next several years.

The talent and security dimensions are particularly significant. They suggest that the bottleneck to scaling AI affordably is not just compute capacity, but the organisational capability to secure systems and staff them properly — factors that are harder to plan for than a straightforward hardware or licensing cost curve.

By the numbers

• 75% — the potential increase in IT costs Bain & Company attributes to AI, even under careful investment scenarios.
• Less than a decade — the timeframe over which Bain projects this cost escalation to occur.

The Renascence take

The instinct in many boardrooms is to treat AI spend as a line item that will shrink once initial build-out is done. Bain's projection challenges that assumption directly, and the reasons why are as much about people and trust as about servers.

Most organisations are budgeting for AI as if it were a one-off technology purchase, when in practice it behaves more like a permanent shift in operating cost structure — closer to how compliance or cybersecurity budgets evolved. The real signal here is that security and talent, not raw compute, are the costs leaders consistently underestimate. A customer-obsessed operator should stress-test its AI business case against rising, not falling, unit costs, and should be explicit with stakeholders that the payoff from AI is more likely to come from better decisions and experiences than from IT budget relief.