Alphabet Inc Class CToken usage surge increases demand for AI services, but pricing challenges create uncertainty.
Goldman Sachs forecasts that token usage, a core cost of artificial intelligence, will increase 24-fold by 2030, reaching 120 septillion tokens per month. Even as per-unit costs decline, the massive surge in usage driven by the adoption of AI agents in enterprises is pushing total expenses higher and making them difficult to predict. Providers such as Microsoft, Google, OpenAI, and Anthropic are facing a major challenge in pricing their services to be both profitable and sustainable, because the true cost depends on the number of tokens used, which cannot be forecast in advance, and frequent price changes prevent corporate customers from budgeting effectively. Experts say that making AI sustainably profitable may be an even tougher challenge than developing powerful AI.
Alphabet Inc Class CToken usage surge increases demand for AI services, but pricing challenges create uncertainty.
Microsoft CorporationMicrosoft faces pricing challenges due to unpredictable token costs, but benefits from increased AI adoption.
Goldman Sachs Group IncOpenAI's profitability is challenged by token cost unpredictability, though demand surges.