Who pays for AI’s $2.6 trillion build-out? Look to corporate budgets
Reported by Fortune1 source ↓
BCG estimates the global AI ecosystem will need $2.6 trillion in annual spending by 2030, with enterprises accounting for $1.8 trillion. The enterprise figure is a measure of required spending, not a forecast.
Key points
- BCG says enterprises would need to supply $1.8 trillion of annual AI spending by 2030
- BCG’s scenario combines $857 billion shifted from existing tech budgets with $937 billion in new spending
- The estimate covers global spending but excludes China
- BCG estimates each gigawatt of capacity coming online in 2026 needs about $15 billion in revenue to cover costs
Details
Fortune reports that a BCG Institute analysis estimates the AI ecosystem will require about $2.6 trillion in annual end-market spending by 2030 to support data centers, chips, model development and investor returns. Alongside $1.8 trillion from enterprises, the estimate includes $366 billion from consumers, $204 billion from hyperscalers’ internal use and $200 billion from emerging technologies such as physical AI.
The estimates cover global spending excluding China, with infrastructure costs concentrated mainly in the U.S. BCG’s enterprise figure is a requirement estimate, not a prediction that companies will spend that amount. In one illustrative scenario, companies would shift $857 billion from existing technology budgets and add $937 billion in new spending, putting enterprise tech budgets about 13% above an IDC IT spending forecast.
BCG says AI’s share of enterprise tech budgets would rise from 5% today to 25% in 2030 in that scenario. BCG’s Clark O’Niell said companies seeking more AI investment should make a case based on productivity gains, cost reductions, new revenue or competitive advantage; he also identified infrastructure financing and revenue generated by deployed compute as indicators to watch.
BCG assumes AI chips last an average of five years and estimates that each gigawatt of capacity coming online in 2026 must generate about $15 billion in revenue to cover its costs. The analysis warns weaker demand could lead to price competition and losses for model providers and data-center operators; O’Niell advises most corporate buyers to avoid long-term price-per-token commitments while prices fall.
Written from coverage as of
Images
- Fortune