Business 4 min read By Arthur Ellington
Enterprises would need to fund most of AI’s $2.6 trillion build-out, BCG says
A BCG Institute analysis estimates that corporate budgets would have to carry the bulk of the AI ecosystem’s $2.6 trillion annual spending need by 2030, raising questions about returns and financing risk.
Corporate budgets would need to carry most of the artificial intelligence industry’s spending by 2030, according to a BCG Institute analysis. The consultancy estimates the AI ecosystem will require about $2.6 trillion in annual end-market spending to support data centres, chips, model development and investor returns, with enterprises accounting for $1.8 trillion of that total.
The remaining sums are smaller by comparison. Consumers would account for $366 billion, hyperscalers’ internal use for $204 billion, and emerging technologies such as physical AI for $200 billion. The estimates cover global spending excluding China and reflect infrastructure costs concentrated mainly in the United States, where most new capacity is being built.
BCG is careful to describe the $1.8 trillion as a measure of what the ecosystem would need from enterprises, not a forecast of what companies will actually spend. In one illustrative scenario, companies could contribute that sum by reallocating $857 billion from existing technology budgets and adding $937 billion in new spending. That would lift enterprise technology budgets roughly 13 per cent above the spending forecast from International Data Corporation, with AI’s share of those budgets rising from 5 per cent today to 25 per cent in 2030.
The scenario highlights how heavily the industry’s growth depends on corporate willingness to spend. For chief financial officers seeking more AI investment, the task is to show boards evidence of productivity gains, cost reductions, new revenue or competitive advantage, said Clark O’Niell, BCG’s global lead for AI, infrastructure and cyber. Some strategic benefits, however, may be hard to capture in traditional return-on-investment calculations.
“The question isn’t just whether AI delivers an immediate financial return,” O’Niell said. “It’s also what the cost of falling behind could be.” Fear of falling behind can be a legitimate strategic consideration, but it should not replace a sound business case in place of rigorous analysis.
BCG considers the AI build-out more resilient than previous infrastructure booms, partly because hyperscalers have financed much of their investment through cash flow rather than debt. The ability to fund infrastructure, however, does not guarantee adequate returns. O’Niell points to two indicators to watch: how infrastructure investment is financed, with a sustained shift toward debt signalling greater risk, and whether deployed computing capacity generates sufficient revenue. As frontier AI labs move into public markets, their disclosures could show whether revenue is keeping pace with compute costs.
The economics leave limited room for disappointment. BCG assumes AI chips have an average five-year lifespan and estimates that each gigawatt of capacity coming online in 2026 must generate about $15 billion in revenue to cover its costs. Weaker-than-expected demand could trigger price wars and losses for model providers and data-centre operators.
For corporate buyers, those pressures have practical implications. O’Niell advises most companies to avoid long-term price-per-token commitments, since AI prices continue to fall. Contract flexibility could help finance chiefs capture future cost reductions. The $2.6 trillion figure is achievable, he said, “but it’s not guaranteed.” Even if enterprises supply the spending, the open question is whether they will capture enough of the economic value to justify it.
Source: fortune.com
9
