OpenAI says the gap between companies that merely experiment with AI and those that actually run it is widening fast. In two reports published August 12, the company found that frontier firms — its term for the top 10% of enterprise customers by monthly AI usage — now generate 8.3× as many output tokens per active user as typical firms, up from 2.6× in January, according to OpenAI’s Enterprise Signals report. That measure is a proxy for depth of use: these organizations hand agents longer, multi-step tasks rather than asking chatbots to draft a quick answer.
The gap is clearest in how aggressively frontier firms wire agents into their tooling. Each week, 21% of active users at frontier firms use Plugins, versus 9% at typical firms, and just 3% use skills — the companion building blocks for these workflows — while inside OpenAI itself 95% of employees use Plugins weekly, OpenAI’s Enterprise Signals report says. Plugins bundle reusable instructions with connections to company data and apps so an agent can pull live CRM or document context and take action, as described on Plugins in ChatGPT.
Software engineering was the early home of agentic work, but the load is spreading quickly. As of June, Codex produced 64% of combined Codex and ChatGPT output tokens among enterprise customers, and since February weekly active enterprise Codex users grew 108× in legal, 41× in sales, 41× in recruiting, and 26× in marketing versus just 5× in engineering, OpenAI’s Enterprise Signals report states. Virgin Atlantic shows what that looks like in practice: engineering teams refactored legacy code in 30 minutes instead of two weeks, while product teams compressed weeks of competitive research into hours while shaping a five-year digital strategy. The shift is also age-skewed in a way surveys miss — six months after adoption, early-career employees sent 13 more messages per week than executives, from a sample of more than 10 million conversations, and among the U.S. public companies in the companion working paper, enterprise adopters held more assets, employed more workers, and invested more in R&D than non-adopters.
OpenAI’s agenda for leaders is to close the frontier gap by connecting agents to company context and tools, setting clear permissions and human review, and turning effective individual workflows into shared practice. For teams building that internal capability, OpenAI Academy’s Three-Course Path for AI at Work maps a route from first use to deployment.
