OpenAI CEO Sam Altman warns AI token costs now exceed human labor expenses

OpenAI CEO Sam Altman has flagged AI token costs as a major operational concern for enterprise clients. Companies are burning through annual software budgets within weeks of starting the fiscal year due to aggressive AI usage patterns. This spending surge is creating immediate pressure on IT departments to justify tooling expenses.

Companies report spending entire annual budgets in Q1 due to aggressive AI usage patterns

The core issue centers on how organizations deploy large language models at scale. Engineering teams and internal developers are running autonomous agents that consume vast amounts of processing tokens. Some users hit 100 billion tokens monthly, a massive jump from the 100k baseline recorded six years ago.

Token pricing structures have not kept pace with this exponential growth in consumption. Running AI workloads now costs more than hiring human staff for equivalent tasks. This cost inversion forces companies to audit their AI tooling and cut back on unverified licenses.

OpenAI is actively refining its models to improve efficiency per token. The company acknowledges that clients are treating budget overruns as a widespread operational meme. Internal feedback loops show executives demanding faster ROI from every dollar spent on API access.

The broader industry is grappling with these financial realities. Competitors like Microsoft have already restricted certain AI tool licenses after cost spikes. Uber leadership notes no proven correlation yet between heavy spending and successful product delivery, highlighting the need for measured adoption strategies.