Hello and welcome to Eye on AI. In this edition….Insights from Fortune Brainstorm AI San Francisco…Disney invests $1 billion in OpenAI and licenses its IP to the company…OpenAI debuts GPT-5.2 in effort to silence concerns it’s trailing rivals…Oracle stock takes a tumble.
Hi, it’s Jeremy here. I’m still buzzing from Fortune Brainstorm AI San Francisco, which took place earlier this week. We had a fabulous lineup including Brad Lightcap, OpenAI’s chief operating officer, Google Cloud CEO Thomas Kurian, Intuit CEO Sasan Goodarzi, Exelon CEO Calvin Butler, Databricks CEO Ali Ghodsi, Rivian CEO RJ Scaringe, Insitro CEO Daphne Koller, and many more. We also had a thoughtful conversation on AI’s impacts with actor, director, and increasingly AI thought leader Joseph Gordon Levitt, as well as a scream of a session with actor, comedian and AI CEO Natasha Lyonne. Today, Sharon Goldman, Bea Nolan, and I are going to share a few highlights and personal impressions.
For me, there was a notable vibe this year that a lot of companies are substantially further along in implementing AI across their organizations, including using AI agents in some limited, but important, capacities. Many audience questions, especially in some of the breakout sessions, were around governance and orchestration methods for an increasingly hybrid workforce where AI agents will be completing tasks alongside employees.
Still, it was striking to hear Butler, the Exelon CEO, say that his company is moving cautiously. When the consequence of getting something wrong is literally lights out, security and reliability have to take precedence over everything else. And so Butler said he was happy not to be a “first mover” but instead a “fast follower” when it came to AI implementations. Let other people take the hit and learn from their mistakes, seems to be his view.
And this wasn’t the only place where speakers were seeking to tamp down hype. It was refreshing to hear Michael Truell, the cofounder and CEO of hit coding assistant Cursor tell me that he didn’t think software engineering would ever be fully automated in the way that OpenAI CEO Sam Altman sometimes talks about. Instead Truell said that while the amount of time that coders spent on “compilation” of code would continue to shrink, he saw a continued need for humans to make design decisions around “how should the software work.”
Similarly, Vidya Peters, from DataSnipper, said she thought there would still be a role for qualified accountants within finance organizations, even if they were increasingly being assisted with AI tools such as the one her company makes. She also said she thought that applications geared specifically for a particular industry or job—especially in regulated industries—would continue to win out over more general purpose AI models, even as the big AI companies are increasingly targeting specific professional use cases for their general purpose models.
A panel that Sharon moderated on the “new geography of data centers” was fascinating. The message was that right now, data centers are going where the power is. But increasingly data centers are going to be looking to build their own power on site and possibly even become net contributors to the grid. And Jason Eichenholz, the CEO of Relativity Networks, said that as AI inference workloads come to eclipse AI training workloads, there will be an increasing need to bring data centers close to major population centers, but that most cities in the U.S. are power constrained. How are we going to get these urban centers the tokens they need at the speed at which they need them? That’s anyone’s guess right now, Eichenholz says—although his company builds the fast fiber that will carry those tokens from the data centers to end users.
Finally, I enjoyed hearing Dayle Stevens from Telstra explain why her company chose to form a joint venture with Accenture to deliver its AI stragegy, rather than simply hiring the consulting firm under a traditional service contract. Stevens said the joint venture has enabled the company to move much faster than it would have otherwise and to tap expertise, including starting an AI innovation hub in Silicon Valley, that would have been hard to implement otherwise.
The future of enterprise AI is hybrid
Now, here’s Sharon’s takeaways: In my mainstage session with PayPal global head of AI Prakhar Mehrotra and Marc Hamilton, VP of solutions architecture and engineering at Nvidia, both discussed the increasing power of open source AI models to allow enterprise companies to control their data and fine-tune for specific use cases. But both agreed that the future of enterprise AI will be hybrid, with enterprises typically using both open models and proprietary model APIs.
There was plenty of time for philosophizing, as well: at one dinner, I chatted with delegates from The Clorox Company, Workday and other companies about everything from what jobs were future-proof (I suggested dog walkers were safe from AI) to what AI would really mean for the future of today’s children (the bottom line: they still need to learn to think for themselves!).
My favorite panel was one I moderated with a half-dozen leaders and stakeholders in the world of AI data centers, including Andy Hock from Cerebras, Matt Field from Crusoe, and former OpenAI infrastructure policy leader Lane Dilg. We dug into how the line is blurring between power infrastructure and data centers, with billions in capital and gigawatts of power at play. My biggest takeaway was that the AI data center issue is local, local, local. Every community and local government will be dealing with its own specific issues and compromises around issues such as land, energy, and water—and what works for one area might not work for another.
People and culture are paramount
And here is what Bea had to say about this year’s Brainstorm AI San Francisco:
Most enterprises are still trying to figure out the best way to adopt AI, but leaders this year were also keen to emphasize that choosing the right tools is only part of the equation. Companies also need to ensure that both their employees and their org charts are ready for the shift—otherwise, even the most advanced AI pilots are likely to fail.
As Accenture’s Chief Responsible AI Officer Arnab Chakraborty put it: “Don’t just think about technology—think about people and the culture. It is so paramount.”
Or take Open Machine CEO Allie K. Miller’s advice and don’t call AI a tool at all: “Calling it a tool ends up being a little bit of borderline self-limiting behavior that is holding enterprise all around the world behind.”
I also moderated a panel of healthcare experts, which brought together a mix of clinicians who see patients every day and tech leaders building and deploying healthtech tools at scale. In healthcare, the industry is generally feeling good about clinician-facing AI, but it’s still wrestling with what it means to safely deploy patient-facing agents.
The panelists discussed, among other things, what it means to be moving toward a future where patients and clinicians consult the same AI before they consult each other.
The excitement is running high on the corporate side, but not that much has really changed in the examination room—at least according to Gurpreet Dhaliwal, a clinician-educator and Professor of Medicine at the University of California. Whether it’s with Dr. Google, Dr. ChatGPT, or just a neighbor with some strong beliefs about antibiotics, Dhaliwal said patients have always arrived with a second opinion in their back pocket. While AI is poised to be a revolutionary force for healthcare—especially in fringe cases such as rare diseases—it’s yet to fundamentally change the dynamic between patients and their physicians.
With that, here’s the rest of the AI news.
Jeremy Kahn
jeremy.kahn@fortune.com
@jeremyakahn
FORTUNE ON AI
Google DeepMind agrees to sweeping partnership with U.K. government focused on science and clean energy—by Jeremy Kahn
Hinge’s founder and CEO is stepping down to start a new AI-first dating app—by Marco Quiroz-Gutierrez
Cursor has growing revenue and a $29 billion valuation—but CEO Michael Truell isn’t thinking about an IPO—by Beatrice Nolan
AI IN THE NEWS
Disney invests $1 billion in OpenAI, brings characters to OpenAI apps. The home of Mickey Mouse is investing $1 billion in OpenAI and, under a three-year licensing deal, will let users generate short, prompt-driven videos in OpenAI’s Sora app using more than 200 Disney, Marvel, Star Wars, and Pixar characters. OpenAI is supposed to create guardrails to prevent users from creating videos or images that might reflect poorly on the Disney brand. The partnership was struck after nearly two years of talks. Meanwhile, Disney simultaneously sent a cease-and-desist letter to Google accusing it of large-scale copyright infringement tied to AI outputs featuring Disney characters. You can read more from The Wall Street Journal here.
OpenAI debuts GPT-5.2 model, answering concerns it was trailing competitors. The company launched a new AI model that, according to evaluations OpenAI conducted, delivers state-of-the-art performance across a wide range of tasks, including coding, mathematical reasoning, and “knowledge work.” The model showed significant improvement over GPT-5.1, which OpenAI released only a month ago, and bested Google’s and Anthropic’s new models. The release of Google’s Gemini 3 Pro in late November prompted OpenAI CEO Sam Altman to declare a “code red” to refocus the company on improving ChatGPT. But OpenAI executives said the release of GPT-5.2 had been in the works for months and that its debut was not related to the “code red.” OpenAI said GPT-5.2 also improves safety, particularly around mental health–related responses. You can read more from Jeremy here.
New lawsuit claims ChatGPT contributed to murder-suicide in Connecticut. A wrongful-death lawsuit was filed against OpenAI and Microsoft after a 56-year-old Connecticut man, Stein-Erik Soelberg, killed his 83-year-old mother and then himself following months of increasingly delusional conversations with ChatGPT. His family says the chatbot reinforced and contributed to his mental illness. OpenAI has expressed condolences and pointed to ongoing improvements to ChatGPT’s ability to recognize and respond to users in distress. You can read more from The Wall Street Journal here.
Microsoft says health queries are the most frequent use of its Copilot AI by consumers. Microsoft analyzed 37.5 million anonymized Copilot conversations from January through September 2025 to understand how people use the AI assistant in daily life. The study found that health-related questions dominated mobile usage, while topics and usage patterns varied significantly by device, time of day, and context. Beyond information search, users increasingly turned to Copilot for advice on personal topics, showing its role as a companion in both work and life moments. You can read Microsoft’s blog on the findings here.
Meta and Eleven Labs sign a new partnership to provide voice overs for Reels. Meta has partnered with London-based voice AI company ElevenLabs to integrate AI-powered audio capabilities across Instagram and Horizon. This partnership will enable new features such as the ability to dub Reels into local languages as well as to generate character voices. You can read more in The Economic Times here.
AI CALENDAR
Jan. 7-10: Consumer Electronics Show, Las Vegas.
March 12-18: SWSW, Austin.
March 16-19: Nvidia GTC, San Jose.
April 6-9: HumanX, San Francisco.
EYE ON AI NUMBERS
$34 billion
That's the one-day paper loss Oracle founder and chairman Larry Ellison suffered Thursday after his company’s shares were pummeled by investors increasingly concerned with the amount Oracle is spending to build data centers for OpenAI. Oracle’s quarterly capital expenditures for the last quarter came in above analyst expectations and in fact exceeded the amount of cash the company generated in the quarter. “It’s like the poster child of the AI bear case,” Jay Hatfield, chief executive of Infrastructure Capital Advisors, told the Wall Street Journal.
