OpenAI studies show frontier firms use 8.3x more AI tokens, shift to agentic execution
OpenAI published two studies on enterprise AI adoption, finding that frontier firms—the top 10% of monthly AI users—generate 8.3 times more output tokens per active user than typical firms. The research highlights a shift from assistance to execution, with agentic AI connecting to company context, tools, and workflows. The studies, 'Enterprise Signals' and 'How Organizations Use AI: Evidence from ChatGPT', examine adoption across companies, roles, and seniority levels.
Coverage timeline
OpenAI Blog
Organizations are expanding both where they use AI and what they ask it to do. Enterprise AI is moving from asking to doing, yet not all firms are making that transition at the same pace. Frontier firms—those in the top 10% of AI usage each month—now generate 8.3× as many output tokens per active user as typical firms. The measure is a proxy for depth of use, and the widening gap appears alongside greater adoption of capabilities that connect agents to company context, tools, and repeatable workflows. Today we are publishing two complementary studies that examine this shift. Enterprise Signals leads with a practical view of agentic AI across OpenAI’s enterprise customer base, including what frontier firms are doing differently and where agentic work is spreading. The companion working paper, How Organizations Use AI: Evidence from ChatGPT(opens in a new window), examines how adoption grows across companies, roles, and levels of seniority. Together, the studies point to a practical ent