Bain: AI industry must reach $6T annual revenue by 2031 to justify data center capex

Bain's framing turns the data-center buildout into a revenue-underwriting problem rather than a capacity story. The gap it identifies is not a forecast of failure but a map of where the burden of proof now sits: consumer and enterprise AI services alone cannot carry the capex, so the thesis depends on categories that are still largely pre-commercial. That shifts the risk from whether compute gets built to whether the applications Bain lists scale fast enough to absorb it. Whether those emerging areas attract the investment and adoption needed to close the gap is the open question.