--- title: OpenAI's own data shows the AI rich are getting richer description: OpenAI's enterprise adoption reports reveal a widening usage gap: top-10% enterprise users now generate 8.3x more output tokens per active user than typical... canonical: https://daily.dev/posts/openai-s-own-data-shows-the-ai-rich-are-getting-richer-a6mwlkptp twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: OpenAI's own data shows the AI rich are getting richer | daily.dev og:description: OpenAI's enterprise adoption reports reveal a widening usage gap: top-10% enterprise users now generate 8.3x more output tokens per active user than typical... og:url: https://daily.dev/posts/openai-s-own-data-shows-the-ai-rich-are-getting-richer-a6mwlkptp og:image: https://api.daily.dev/og/posts/A6MwlkPtP.png og:image:alt: OpenAI's own data shows the AI rich are getting richer og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # OpenAI's own data shows the AI rich are getting richer **[Trends](https://daily.dev/sources/trends)** · 3 min read · 2 upvotes · 0 comments ## Summary OpenAI's enterprise adoption reports reveal a widening usage gap: top-10% enterprise users now generate 8.3x more output tokens per active user than typical firms, up from 2.6x in January. Frontier firms use plugins and skills at far higher rates (21% vs 9%, 19% vs 3%). Ethan Mollick notes these leading firms already had the most productive employees, suggesting AI compounds existing advantages rather than leveling the field. Separately, Codex now makes up 64% of combined Codex-plus-ChatGPT output tokens, with adoption surging outside engineering in legal, sales, and marketing. Early-career employees now send more messages weekly than executives, suggesting habit matters more than seniority or access. ## Content OpenAI dropped some enterprise adoption numbers this week, and the takeaway isn't flattering for most companies. There's a small group of firms pulling dramatically ahead on AI usage, and the gap is widening, not closing. The headline stat: the top 10% of enterprises use plugins twice as often as typical firms (21% vs 9% weekly) and skills six times as often (19% vs 3%). OpenAI's framing is blunt: "these frontier firms are not ahead by accident." They've built the habit of reaching for advanced capabilities while everyone else pokes at the basics. But the more striking number is the trend line. Frontier firms now generate 8.3x as many output tokens per active user as typical firms, up from 2.6x back in January. That's not a gap holding steady. That's a gap accelerating in real time. Codex adoption tells a similar story about where the work is actually going. It now accounts for 64% of combined Codex and ChatGPT output tokens among enterprise customers, meaning companies are increasingly delegating real execution to agents rather than using AI as a sidekick. And it's not staying in engineering: weekly active Codex users grew 108x in legal, 41x in sales and recruiting, and 26x in marketing since February, compared to just 5x in engineering itself. The report also flags that early-career employees are out-messaging executives six months into adoption, which suggests seniority isn't what determines who benefits. Ramp's spending data, pulled from a16z's numbers, makes the disparity almost absurd. Median companies spend $12 per employee per month on AI. The top 1% spend $7,500. As @rohanpaul_ai put it, "the top 1% burns a median company's entire annual AI budget in a single afternoon." Same tools, same two years on the market, 625x apart in spend. Academic Ethan Mollick's read on all this is the uncomfortable part: "early AI adopting firms that were already doing well may start to outpace others." In other words, this might not be a story about AI leveling the playing field. It might be a story about AI compounding whatever advantages already existed. The tools are the same for everyone, sure, but access was never the bottleneck. Usage is, and usage habits don't spread evenly. If the median firm doesn't need to catch up to the top 1%, just move a little closer, that's still enough to multiply token demand across a much bigger base over the next few years. Worth watching whether "everyone has the same tools" turns out to be cold comfort or just a footnote before the gap gets permanent. ## Questions this post answers ### How much more do frontier enterprises use AI compared to typical companies according to OpenAI's data? Frontier enterprises generate 8.3 times as many output tokens per active user as typical firms, up sharply from 2.6 times back in January. The top 10% of enterprises also use plugins twice as often (21% vs 9% weekly) and skills six times as often (19% vs 3%) as typical firms, showing a widening rather than closing gap. _Track how AI usage gaps between companies evolve by following enterprise AI adoption coverage on daily.dev._ ### How does Codex adoption compare to ChatGPT usage among OpenAI's enterprise customers? Codex now accounts for 64% of combined Codex and ChatGPT output tokens among enterprise customers, indicating companies increasingly delegate real execution to agents rather than treating AI as a supplementary tool. Adoption is also spreading fast outside engineering, with weekly active Codex users growing 108x in legal, 41x in sales and recruiting, and 26x in marketing since February, versus only 5x in engineering. _Developers weighing agent-based tools like Codex against chat assistants can follow adoption trends on daily.dev._ ### How much do top-spending companies pay for AI tools compared to median companies? The top 1% of companies spend a median of $7,500 per employee per month on AI, compared to just $12 per employee per month at median companies, a 625x gap despite using the same tools available for roughly the same two years. This spending disparity, drawn from Ramp data via a16z, illustrates that access to AI tools was never the bottleneck; usage habits were. _Anyone benchmarking AI budget decisions against peers can find related spending analysis on daily.dev._ ## Community take How the wider developer community reacted, aggregated from 1 discussion and 5 comments across x (as of 2026-09-13). **TL;DR:** Reactions are minimal and mostly light, informal responses to the widening usage gap, with one comment noting how stark the divide feels rather than substantive debate. **Sentiment:** 10% positive · 80% mixed · 10% skeptical **The pushback** - The gap between heavy and typical AI users feels like an entirely different game, stretching budgets even with small tweaks. **By community** - x (mixed): A handful of brief, low-substance replies with light humor and one comment remarking on the stark usage gap. **Highlights** > @rohanpaul_ai @rohanpaul_ai wild gap, right? feels like a different game entirely. even small ai tweaks stretch budgets here. > — [itsthedonhashim on x · 2 points](https://x.com/itsthedonhashim/status/2089078809928552766) **Source threads** - [x](https://x.com/rohanpaul_ai/status/2089058029652336716) · 0 points · 5 comments ## Similar posts on daily.dev - [The most AI-obsessed companies spend $7,500 per employee per month. The median spends $11.](https://daily.dev/posts/the-most-ai-obsessed-companies-spend-7-500-per-employee-per-month-the-median-spends-11--iomhws4gs) · The Next Web · 0 upvotes · 0 comments --- Tags: [#openai](https://daily.dev/tags/openai), [#chatgpt](https://daily.dev/tags/chatgpt), [#openai-codex](https://daily.dev/tags/openai-codex) [View this post on daily.dev](https://daily.dev/posts/openai-s-own-data-shows-the-ai-rich-are-getting-richer-a6mwlkptp) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"OpenAI's own data shows the AI rich are getting richer","url":"https://daily.dev/posts/openai-s-own-data-shows-the-ai-rich-are-getting-richer-a6mwlkptp","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/openai-s-own-data-shows-the-ai-rich-are-getting-richer-a6mwlkptp"},"datePublished":"2026-08-13T23:18:06.009Z","dateModified":"2026-08-16T21:35:03.382Z","description":"OpenAI's enterprise adoption reports reveal a widening usage gap: top-10% enterprise users now generate 8.3x more output tokens per active user than typical...","image":"https://pbs.twimg.com/media/HPo3b8XagAEDHT0.jpg","thumbnailUrl":"https://pbs.twimg.com/media/HPo3b8XagAEDHT0.jpg","isAccessibleForFree":true,"articleSection":"Trends","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Trends","logo":"https://media.daily.dev/image/upload/s--ZfSp3asX--/f_auto,q_auto/v1780996004/logos/trends?_a=BAMAMiWQ0","url":"https://daily.dev/sources/trends"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/openai-s-own-data-shows-the-ai-rich-are-getting-richer-a6mwlkptp","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":2},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"openai,chatgpt,openai-codex","timeRequired":"PT3M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Trends","item":"https://daily.dev/sources/trends"},{"@type":"ListItem","position":3,"name":"OpenAI's own data shows the AI rich are getting richer"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/openai-s-own-data-shows-the-ai-rich-are-getting-richer-a6mwlkptp#faq","mainEntity":[{"@type":"Question","name":"How much more AI usage do top enterprise firms have compared to average firms?","acceptedAnswer":{"@type":"Answer","text":"Top-10% enterprise users generate 8.3 times as many output tokens per active user as typical firms, according to OpenAI's own enterprise adoption data. That multiple was only 2.6x back in January, showing the usage gap between frontier firms and average firms is widening rather than closing over time. daily.dev helps teams track how AI adoption gaps between companies are evolving."}},{"@type":"Question","name":"What percentage of ChatGPT enterprise output tokens now come from Codex?","acceptedAnswer":{"@type":"Answer","text":"Codex accounts for 64% of combined Codex-plus-ChatGPT output tokens among enterprise customers, indicating a shift from asking AI for help toward letting it directly execute tasks. Codex weekly active users grew 108x in legal, 41x in sales and recruiting, and 26x in marketing since February, while engineering only grew 5x. Teams weighing AI-assisted coding versus autonomous execution can follow shifts like this on daily.dev."}},{"@type":"Question","name":"Do senior employees use AI tools more than junior employees at work?","acceptedAnswer":{"@type":"Answer","text":"No, six months after adoption, early-career employees send more messages per week using AI tools than executives do. Seniority does not appear to confer an AI usage advantage; habitual use of advanced capabilities matters more than rank or access to the tools. daily.dev keeps developers informed on how AI usage patterns differ across seniority levels."}}]} ```