OpenAI, Anthropic, Google and other large AI labs are collectively losing hundreds of millions of dollars each month by deliberately under-pricing enterprise AI subscriptions, according to an 11 May analysis on thestateofbrand.com.
The labs are pursuing a coordinated loss-leader strategy that charges corporate customers “gas-station hot-dog prices” for capabilities that cost far more to deliver, thestateofbrand.com reports. Rather than a temporary discount, the pricing gap is described as a deliberate “gulf” baked into current subscription tiers. No single company is named as the architect; instead, the piece frames the practice as an industry-wide program without precedent in scale.
The arrangement places every enterprise that has embedded AI into workflows atop what thestateofbrand.com labels a “ticking time bomb.” Once the labs pivot to sustainable pricing, customers face either sharp cost increases or disruptive service cuts. Comparable subsidy phases in cloud computing (AWS’s early 2000s losses) and ride-hailing (Uber’s decade-long burn) suggest eventual repricing is inevitable, but none reached the current burn rate relative to revenue now seen in generative AI.
Enterprises should expect the first repricing moves within the next 12–18 months, thestateofbrand.com warns. Watch for new metered-usage clauses, seat minimums or outright tier eliminations in upcoming contract renewals. Procurement teams that lock multi-year deals before such changes take effect may be the only near-term hedge against the looming bill.