According to HSBC Global Investment Research, the rapid rise in artificial intelligence (AI) investments by hyperscalers is set to capture a significant portion of enterprise IT spending, potentially squeezing the traditional software services outsourcing industry in the near term.
The report estimated that this impact will become visible over the next three to four years as contracts come up for renewal, with an annual effect of 3-4 per cent projected during 2025-27. So far, IT companies have offset this pressure through higher client volumes, but the report expects 2026 to bring a “push and pull” between deflationary pressures from AI adoption and macroeconomic tailwinds.
As per the report, capital expenditure by the top four hyperscalers is expected to reach about $300 billion in 2025 and rise to $500 billion by 2030. This level of spending could absorb the majority of incremental enterprise technology budgets.
Last year, Amazon Web Services (AWS), Microsoft Azure, and Google Cloud Platform added nearly $46 billion in revenue, bringing their combined cloud revenue to about $260 billion. In comparison, the IT services industry’s revenue grew by only $14 billion, or just over 5 per cent, to $283 billion.
The report mentioned that the consolidation of enterprise IT spending toward hyperscalers and AI-driven automation will significantly impact traditional IT service providers.
In one of the largest recent cloud deals, OpenAI signed a $300 billion contract with Oracle to secure computing power over five years. The agreement positions Oracle as a new challenger to the above mentioned hyperscalers while underscoring the scale of investment flowing into AI data centres.
Despite near-term challenges, the report expects the IT services sector to grow 5–6 per cent in 2026-27, with work volumes expanding 8-10 per cent. The report noted that, much like the earlier shift from on-premise software to software as-a-service (SaaS), the AI transition will initially benefit IT firms that guide clients through this change, though many business processes are expected to eventually become autonomous.
