OpenAI released fresh data on Monday showing a dramatic surge in how enterprises are using its AI tools. ChatGPT message volume has grown eightfold since November 2024, and employees report saving as much as an hour each day. This comes about a week after CEO Sam Altman circulated an internal “code red” memo warning about Google’s competitive threat.
The timing emphasizes OpenAI’s effort to position itself as the leading enterprise AI provider, even as it faces growing pressure. Roughly 36% of U.S. businesses use ChatGPT Enterprise, compared with 14.3% for Anthropic, according to Ramp AI Index. Yet, most of OpenAI’s revenue still comes from consumer subscriptions, a base that Google’s Gemini could challenge. OpenAI also competes with Anthropic, whose B2B sales are a primary revenue source, and with other open-weight model providers increasingly targeting enterprise customers.
OpenAI has committed about $1.4 trillion in infrastructure investments over the coming years, making enterprise growth crucial to its business model.
From an economic perspective, OpenAI’s chief economist Ronnie Chatterji stresses that consumer use matters for growth, but historically transformative technologies yield the biggest gains when firms adopt and scale them. The new findings suggest larger enterprises aren’t just adopting more; they’re integrating these tools into everyday workflows. Not only are more messages being sent, but organizations using OpenAI’s API are consuming 320 times more “reasoning tokens” than a year ago, indicating growing use for complex problem-solving. This could reflect genuine value or heavy experimentation that may not translate into long-term payoff.
The rise in reasoning tokens is linked to higher energy usage in data centers, which could make the expansion expensive and potentially unsustainable long-term. TechCrunch has requested OpenAI’s take on enterprise budget planning for AI and whether this growth rate is viable.
Beyond raw usage, deployment patterns are shifting. Use of custom GPTs—tools that encode company knowledge into assistants or automate workflows—jumped 19x this year and now accounts for about 20% of enterprise messages. OpenAI highlighted BBVA, a digital bank, as a client regularly using more than 4,000 custom GPTs.
Brad Lightcap, OpenAI’s chief operating officer, notes that this demonstrates how people can tailor powerful technology to practical, business-specific needs. These integrations have translated into meaningful time savings: respondents report saving 40 to 60 minutes per day using OpenAI’s enterprise products, though this figure doesn’t account for time spent learning, prompting, or correcting outputs.
The report also shows that AI is expanding workers’ own capabilities. About three-quarters of surveyed employees say AI enables them to perform tasks they previously couldn’t, including technical work, and there’s a 36% rise in coding-related messages beyond traditional tech and research teams.
While the message underscores democratizing access to skills, there’s a caveat: greater reliance on AI could introduce more security risks. Lightcap pointed to OpenAI’s private-beta agentic security researcher, Aardvark, as a potential tool for detecting bugs and vulnerabilities.
The findings also reveal that even among the most active ChatGPT Enterprise users, many aren’t yet leveraging the most advanced features like data analysis, deep reasoning, or comprehensive search. Lightcap suggests this slow uptake reflects a broader mindset shift and deeper integration with company data and processes. Full adoption will take time as organizations adjust workflows to capitalize on what AI can do.
Both Lightcap and Chatterji highlighted a growing divide in AI adoption: some frontline workers use more tools to save more time, while others lag behind. Lightcap notes that some firms treat these systems as a piece of software to buy, while others are re-platforming large swaths of their operations around AI.
OpenAI casts this as an opportunity for laggards to catch up, though for workers tasked with training AI to imitate their jobs, the pace may feel daunting.
Rebecca Bellan, a senior TechCrunch reporter covering AI business trends, contributed to this report. For inquiries, she can be reached at rebecca.bellan@techcrunch.com or via Signal at rebeccabellan.491.
