Citing an unnamed source, CNBC reported that OpenAI CFO Sarah Friar said at a recent investor meeting that OpenAI “entered the year at 60-40, but enterprise has accelerated much faster than expected and those lines have now crossed … The majority of our revenue is now enterprise.”
AI
Artificial intelligence is more a part of our lives than ever before. While some might call it hype and compare it to NFTs or 3D TVs, generative AI is causing a sea change in nearly every part of the technology industry. OpenAI’s ChatGPT is still the best-known AI chatbot around, but with Google pushing Gemini, Microsoft building Copilot, and Apple adding its Intelligence to Siri, AI is probably going to be in the spotlight for a very long time. At The Verge, we’re exploring what might be possible with AI — and a lot of the bad stuff AI does, too.
- RELATED /

Scribble some nonsense, offer wooden jokes, or make your own requests of a real human playing a fake AI.

On The Vergecast: Instagram’s new logo, Zuck’s AI manifesto, and lots of other bad ideas.
Latest In AI
The newly-announced Gemini 3.7 Flash is rolling out today as an option for AI Pro and AI Ultra subscribers. The updated model is better at following instructions and “understanding your intent, so you get even more helpful responses,” says Search VP of product Robby Stein.
Elon Musk’s SpaceX completed its $60 billion acquisition of the AI coding tool Cursor on Friday. Cursor says it’s going to work on the company’s Grok AI chatbot, which added a new AI agent feature this week.
[X (formerly Twitter)]
A new “take notes” button is rolling out now for paid Workspace accounts that records audio from your mic so Gemini can generate notes for in-person meetings. Just like the virtual version of the feature, it also generates a transcript and “action items” and saves them to a Google Doc in your Drive account.
[Workspace Updates Blog]
The new Gemini 3.7 Flash makes the Spark AI agent “more efficient for knowledge work with improved tool use for Google Workspace apps, delivering improved accuracy and output quality for complex, multi-skill workflows,” Google says. The AI model itself also offers “substantial” improvements in software engineering and web development workflows, according to the company.
Google still hasn’t launched its long overdue 3.5 Pro model.
[Google]
The end of USAID, which Elon Musk fed “into the wood chipper”, will cause an extra 14 million deaths through 2030, including 4.5 million in children under the age of 5. SpaceX shares are $141.10 as of this writing, above their IPO price of $135.
Correction: A previous version of this post referred to the journal article’s estimates as “new.” It was published in 2025.



Higgsfield is using Cully Hill Boys to show you how to prompt up your own movie.

Is the shakeup at Google DeepMind a strategy or a crisis?
According to Twitch chief product officer Mike Minton, that’s the “honest answer” as to why the streaming platform has started training Amazon’s generative AI models on users’ content by default. Twitch users can choose to opt out, but Minton defends the default setting because “almost every content service in the world” is also doing it.
The “Ask YouTube” conversational search experience that YouTube started testing in April is now rolling out to the mobile app. It’s available for signed-in users in the US who are 13 or older, and can be found by tapping the “Ask” option in the YouTube search bar.
We live blogged almost 90 minutes of a show that was, honestly, better than expected. Our video team cut it down to eight.
The company working on deals that would let Siri to lean on publisher content for news and information, The Wall Street Journal reports. Apple has apparently proposed a plan where publishers would get paid when content is actually used.
[The Wall Street Journal]
Anthropic updated its Chrome extension so using the side panel bot now saves chats to your history, where you can continue them outside in the browser in Claude’s desktop, web, or mobile apps.
The Cowork version is available today for Max and Team users, and coming to Pro users “over the coming weeks.”
AI data center projects are continuing to pop up across the US, with frequent opposition from locals concerned about their impact. Here are a few recent articles about the projects:
- Atlanta News First: Lovejoy data center proposal withdrawn after community pushback
- FlagerLive.com: Flagler County Echoes Surging Backlash Against Data Centers with One-Year Construction Ban
- MLive.com: Data-center developer sues Oakland County city
- Ohio Capital Journal: All candidates for Ohio governor have now called for restrictions on data center boom
- Spectrum News 1: Community opposition grows as Louisville weighs data center code changes
The rapper admitted to using AI on his ’80s-tinged $TARFACE project and received a rare zero. Jet was the last to receive the dishonor in 2006 for Shine On. Pitchfork pulls no punches, saying:
$TARFACE is smug and complacent, an amalgam of ’80s synth-pop and disco slapped together by someone who’s seemingly never cared to listen to one synth-pop or disco record in his life … Stream it out of morbid curiosity if you must, but know that even a single listen runs the risk of making it profitable and therefore encouraging the music industry to commit such atrocities again.
[Pitchfork]
Lots of smart internet people have been talking about this Lauren Leek essay on the algorithmic flattening of cities and culture and everything else, and it’s as good as advertised. It also offers a shockingly simple way to fix the problem, which is why I spent the morning reading about a machine learning framework known as multi-armed bandits.
“[J]ust be a more adventurous consumer” is exactly the conclusion that lets the systems off the hook. The exploration has to be mandated at the level of the machine, because that’s where real change can happen.
[Lauren’s data Substack]
Seedance 2.5’s ability to generate longer clips that can be carefully edited without starting over scratch are two of the model’s biggest improvements over Seedance 2.0. But Deadline reports, that the new model also much better at detecting and refusing to comply when it’s been prompted to generate outputs based on copyrighted material.
We’re talking offline Gemini, proactive AI suggestions, better GPS maps, insulin resistance trends, and a $50 price hike.
After initially being tested by “select creators,” Creator Studio is now available for iOS users in the US and Canada. The original Creator Studio page manager experience was shut down in 2023, but Meta describes this revived version as an “AI-powered companion app” that helps creators to grow and engage with their audience.
[How creators are using Creator Studio to experiment and grow]


“Together, the companies will develop entirely new music experiences that reimagine how artists and fans can connect,” according to a press release. “The agreement ensures that BMG artists and songwriters who choose to participate have their rights protected and are compensated for their music, while also settling prior use of BMG recordings and publishing works.”
The two companies are also signing the deal ahead of the launch of Suno’s planned music model developed “in partnership with the music industry.”
Chloé Bakalar departed OpenAI last month, less than a year after joining OpenAI as its head of ethics, according to a report from the Financial Times. A source tells the outlet that there currently isn’t a replacement for Bakalar, who previously served as the chief ethicist at Meta for over three years.


Before Alexa and Siri, there was ELIZA. The first chatbot ever was the creation of a man named Joseph Weizenbaum, who built the bot almost as a magic trick — and then couldn’t believe how many people fell for it. David Berry and Mark Marino, two members of a team dedicated to restoring and understanding ELIZA, explain how Weizenbaum created the virtual conversant, why he came to regret it, and what ELIZA can teach us about life in the AI age.
Watch | Listen | Get ad-free
OpenAI announced that the new app is available in preview on Ubuntu, Debian, and Fedora, and includes ChatGPT along with Work and Codex.


Meta acquired the agentic AI startup for $2 billion last year and has already added some of Manus’ tools into its platforms. But now, Manus has announced that it will “return to operating as an independent company” after China blocked the deal.
[Manus]
Most Popular
- Marvel reveals the new X-Men cast, including Inde Navarrette and Adam Driver
- We’re reaching peak camera with the Sony A7R VI
- I finally found a magnetic phone grip I never want to remove
- Xteink’s tiny e-readers are getting access to free books through Libby
- Disney D23 2026: Everything announced for Star Wars, Marvel, and more

























