Oracle (ORCL) co-CEO says the Co. has still not found a way to make generative AI broadly useful to its own workforce, Business Insider reports

The pattern in past hype cycles of this sort is that the trade rarely turns on a single admission; it turns on whether such comments start to corroborate a broader body of evidence about weak end-user adoption and uncertain returns on AI spend, which is the variable that ultimately governs whether hyperscaler and software capex guidance holds. The distinction worth drawing is between internal-use commentary, which speaks to productivity gains, and commentary on customer bookings and backlog, which is what actually feeds the revenue line; the former colours sentiment, the latter moves estimates. Attribution via a media report rather than prepared remarks also matters, as off-script candour of this kind has historically been walked back or contextualised within days, and any such clarification is the immediate follow-on. Worth watching is whether peers or customers echo the adoption difficulty in upcoming earnings commentary, and whether sell-side notes begin separating AI infrastructure demand from evidence of AI software monetisation. As a single executive remark, the signal is about narrative risk to the AI capex theme rather than fundamentals.