Wall Street stopped asking how much AI costs and started asking what it earns back

Bloomberg reported this week that after a brutal stretch for chip and tech stocks, the market’s patience with unexplained AI budgets is running thin. The big four hyperscalers have guided toward as much as $725 billion in capital spending this year, and the question traders are now pressing is no longer how aggressively these companies are investing but what measurable return that investment is throwing off. In 2024 and 2025 the spending itself was rewarded; that reflex has flipped.

The telling detail is how little hard evidence exists to answer the question. Companies have been forthcoming about how much they are pouring into AI and notably quiet about how much revenue it is generating in return - and investors have started to treat that silence as a red flag rather than a rounding error.

The accountability gap the market is punishing at the index level is the same one Uprovd closes at the operating level. Boards and CFOs facing the identical question internally need spend tied to outcomes they can actually point to, not a capital figure with no return beside it. When the public markets start demanding receipts, the pressure to instrument AI value flows straight down to every finance team.