Enterprise Signals

Enterprise Signals is a report from OpenAI that shows, in numbers, how deeply different companies are using AI. The core metric is how many output tokens a single active user generates in a month, tracked as an index over time, used as a gauge for how much heavy-duty work a company is entrusting to AI.

Think of it like a school report card. What matters isn't the raw test score itself, but how much it's improved compared to last term, and whether the gap between top students and average students is widening. OpenAI expresses this gap as a multiple between "frontier" companies—the top 10% by usage—and everyone else. The bigger this multiple gets, the wider the divide between companies that just leave AI switched on as a tool and companies that have handed over their actual workflows to AI.

OpenAI frames this gap not as a simple difference in usage volume, but as a difference in operating approach: how deeply a company connects its own context and tools into agents, and how much substantive work it hands over for AI to repeat reliably.

Articles cite this metric with headlines like "OpenAI reveals token gap between frontier companies has widened to 8.3x," using the numbers to quantify the AI usage gap between companies. One common misunderstanding: this isn't a product a specific company purchases—it's a statistical report OpenAI regularly publishes based on aggregated usage data from its own customers.