Welcome to Eye on AI. Here’s Emily Forlini filling in for Jeremy one last time as his vacation comes to an end. In today’s edition:
- OpenAI doesn’t want to see juicy details in its new report
- Reportedly anthropic Plans a $2 trillion IPO in October – the largest ever
- OpenAI replaces its chief revenue officer after less than a year
- Google declares Sam Altman dead – for 41 minutes
This week I really realized how much money is flowing in the AI industry.
I spoke to two former OpenAI employees who made about $10 million in one day by selling shares in an internal takeover bid Bloomberg Reports In total, all employees amounted to $7 billion. Then, this morning, move on AssetsIn the weekly AI podcast, my colleague Beatrice Nolan and I interviewed the CEO of Lovable. This week, the three-year-old Stockholm-based company doubled its value to $13.3 billion.
A few million, a hundred billion, a trillion (or two, in Anthropic’s case). upcoming IPO) – what’s the difference at this point? There is just one big problem looming in the background: the ROI of AI implementation is still not clear for companies.
OpenAI deals with this existential question in a 69 page report published August 11th about adopting ChatGPT in enterprises. At first glance, the report tells the story of exponential growth in AI usage across all seniority levels and job functions, highlighting what it calls a “frontier gap” in which companies that use AI outpace those that don’t. In other words, if you don’t use AI – especially agents you can delegate tasks to – you lose.
But the fine print tells a different story.
More AI does not mean more money
In a small table on page 35, the researchers report no statistically significant correlation between revenue per employee and those employees’ use of AI, measured in messages sent and tokens used.
“Revenue per employee is not meaningfully linked to issuance tokens per employee or messages per active user when other controls are included,” the report explains.
It says that companies that already have higher revenue per employee tend to be among the first adopters of ChatGPT. Additionally, companies that use the technology more often tend to have higher revenue per employee. In other words, it’s more likely that large, lucrative companies have jumped on the AI bandwagon. However, the study does not clearly show that the more AI they use, the more money they make.
Managers use AI the least
Executives may not be able to measure ROI because they use it the least – another point buried in the report. It’s not just that companies have fewer managers than general employees. However, the interesting thing about the graphic on page 29 is that most senior employees use it less intensivewith the fewest weekly messages per user.
Entry-level professionals are by far the most used, a point that OpenAI CFO Sarah Friar highlighted in her LinkedIn post about the report: “For leaders, this is a reminder that competitive advantage comes from the people closest to the work. Listen to them, learn from them, and help the rest of the organization catch up.”
OpenAI’s corporate sales had a poor result in the fourth quarter of 2025
Surprisingly, the overall usage of OpenAI in enterprises completely flattened from around October 2025 to December 2025. In a graph (page 26) depicting output token growth, the black line representing “overall” growth is almost completely flat for this period. During this time, Anthropic’s Claude Code took the corporate world by storm and became the platform of choice in many places.
To the company’s credit, the line turns upward into an exponential curve in January 2026. A VC told me yesterday, “OpenAI’s execution rate in 2026 was pretty incredible.” OpenAI attributes the growth not only to the addition of new customers, but also to its current customers increasing their usage. We also know that CEO Sam Altman reorganized the company with a view to corporate sales, wiping out what the company called a “business.”Side quests“, such as the video app Sora.
In a sprint to speed up that line – or perhaps to get it moving again? Who knows, the graph ends in March 2026 – OpenAI announced it today rented a new Chief Revenue Officer, Dali Rajic, who will replace Denise Dresser. It’s an aggressive move; Dresser held the role for less than a year. Rajic’s focus will be on accelerating customer adoption and helping companies measure impact as the company moves toward its IPO.
OpenAI paid the scientists who contributed to the report
Finally, two of the five authors are academics who OpenAI paid to support the report. The other three are OpenAI employees. The involvement of academics in work like this usually implies greater credibility and the impartiality of an external research body, but here the situation is somewhat more unclear.
On the first page, David Holtz and Prasanna Tambe are affiliated with Columbia Business School and Wharton at the University of Pennsylvania, respectively. However, a footnote states that both “contributed to this work in their capacity as paid contractors for OpenAI.”
Did the researchers find more that they didn’t publish, as they normally would in a scientific paper? We’ll never know, but it’s another reminder of what has always been true: You need to measure the impact of AI based on your own first-hand experience – not the hype.
Find more AI news here.
Emily Forlini
emily.forlini@fortune.com
@emilyforlini
LUCKY ON AI
Microsoft begins merging consumer and enterprise Copilot apps to create a super app — by Sebastian Herrera
CoreWeave CEO Michael Intrator cites “sold out” capacity as revenue more than doubles and backlog climbs to $104 billion — by Amanda Gerut
When OpenAI employees have a problem, they send an email to this specific address to see if Sam Altman will resolve the issue immediately — by Emily Forlini
CIOs and CTOs have spent years praising AI. Now that costs are rising, they are limiting usage — by John Kell
Forget the gala, these Silicon Valley schools run their own venture capital funds — by Amanda Gerut
AI IN THE NEWS
Anthropic is planning a $2 trillion IPO, possibly in October. It would be the largest company ever recorded and would dwarf SpaceX, but the valuation and timing of the IPO have not yet been determined. Investors expect the AI lab’s annual revenue to be between $100 billion and $120 billion by the end of the year. More inside The Financial Times.
Lovable raises $400 million, doubling valuation to $13.3 billion. The Vibe coding startup is only three years old. The company plans to use the funds for expansion, including increasing its workforce and expanding its Latin American customer base. There are also plans to introduce strict security measures. More inside The Wall Street Journal.
OpenAI’s chief revenue officer is leaving after less than a year. The company has stopped Dali Rajic replaces Denise Dresserwho says she left of her own accord to pursue other endeavors. Many interpret the move as a result of co-founder and president Greg Brockman’s growing presence as the company prepares for a possible initial public offering. More from OpenAI.
AN EYE ON AI NUMBERS
41
Lifespan of OpenAI CEO Sam Altman, according to Google’s botched algorithm. On Wednesday, The search giant declared Altman dead. It added the death date of August 12, 2026 to the fact field displayed at the top of search results. The same panel said Altman is still the current CEO of OpenAI. It’s hard to do anything from the grave, but we’ve seen crazier things in AI. Perhaps it’s a foreshadowing for future CEOs who carry on through AI avatar reproductions of themselves.
Apparently someone had destroyed Altman’s Wikipedia page and claimed he was murdered in Seattle. Accordingly, the perpetrator rewrote the entire introduction to Altman’s Wiki entry in the past tense Gadget Review. Wikipedia’s automated tools flagged the change and a human editor corrected it 41 minutes later – one minute for each of Altman’s supposed years of life.
Google replied on It’s a silly but stark reminder of how fragile the Internet’s information ecosystem is. Some accuse Wikipedia of allowing fraudulent editors, while others say Google is responsible for displaying inaccurate information. It’s unclear whether Google would have caught the problem on its own in 41 minutes or less.
AI CALENDAR
16th-17th November: Fortune 500 Innovation Forum, Detroit. Apply Here participate.
6th–12th December: Conference on Neural Information Processing Systems (Neurips). Sydney, Australia.
7th-8th December: Fortune Brainstorm AI, San Francisco. Apply Here participate.