12 – Breaking News & Latest Updates 2026
Skip to main content

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 /
Robert Hart
Robert Hart
Oura adds a model designed to discuss women’s health to its AI chatbot.

The Oura Advisor chatbot will soon be able to offer smart ring wearers an AI model that it says covers “the full reproductive health spectrum, from early menstrual cycles through menopause.” Of course, reproductive health data is sensitive, particularly in places like the US — you might want to think carefully before handing it over.

Here’s what Oura is saying about the model’s privacy:

It is hosted entirely on Oura-controlled infrastructure, and conversations are never sold, shared, or used to train public or third-party AI systems.

Jess Weatherbed
Jess Weatherbed
AMD forges $100 billion deal with Meta for AI chips.

After buying millions of Nvidia’s AI chips last week, Meta has now also signed a multi-year agreement to buy six gigawatts worth of AMD processors for AI data centers. The deal could see Meta owning 10 percent of AMD’s stock, and follows a similar agreement between AMD and OpenAI.

Dominic Preston
Dominic Preston
Who keeps the safety researchers safe?

Look, it’s never nice to laugh at someone’s misfortune. But when a Meta AI safety researcher puts on a public demonstration of AI safety risks using their personal email as collateral, I think we should all take a moment to enjoy it.

Dkfkhfkwkdnc:

Hang this in the museum. An AI safety researcher pleading with a robot known for being unsafe to not delete her inbox. Sign of many wonderful things to come.

Get the day’s best comment and more in my free newsletter, The Verge Daily.

Inside Anthropic’s existential negotiations with the Pentagon

It’s more than just a $200 million military contract at stake.

Tina Nguyen and Hayden Field
Richard Lawler
Richard Lawler
“STOP OPENCLAW.”

That’s the WhatsApp message Meta safety and alignment researcher Summer Yue sent her AI agent while “watching it speedrun deleting [her] inbox.” Despite the various reasons not to connect OpenClaw to your actual data, apparently, she felt confident based on its results with her toy inbox and moved it to her actual Gmail.

Unfortunately, the AI bot “lost” her instruction not to take action without checking first.

Screenshot of WhatsApp conversation with AI bot reporting it will nuke all old stuff as user messages “STOP OPENCLAW”
Image: Summer Yue (X)
Emma Roth
Emma Roth
OpenAI’s Stargate struggles.

The Information reports on OpenAI’s rocky $500 billion partnership with Oracle and SoftBank, which “has not staffed up and is not developing any of OpenAI’s data centers.”

OpenAI reportedly explored building its own data centers as the three companies disagreed over their roles, but the high costs have led the AI giant to strike individual deals with SoftBank and Oracle instead.

Does Big Tech actually care about fighting AI slop?

It’s harder to clean up a mess you’re still actively making.

Jess Weatherbed
Jess Weatherbed
Jess Weatherbed
X is working on ‘Made with AI’ labels.

The in-development feature reported by @nima_owji shows a toggle that allows users to disclose synthetically made or manipulated content. The move may be in response to India ordering social platforms to embrace provenance systems like C2PA, a standard that X abandoned after Elon Musk stepped in.

A screenshot of X’s in-development ai labels.
These disclosure systems only work if every X user is honest about the images and video content they upload. Which is unlikely.
Image: @nima_owji
Why is AI so bad at reading PDFs?

PDFs are notoriously difficult for machines to parse, in part, because they were never meant to be read by them.

Josh Dzieza
Elizabeth Lopatto
Elizabeth Lopatto
I love a good AI rant.

“If your best idea for what AI can do in the workspace is ‘replace a hundred human beings with a server rack doing the same thing’, you’ve got no business calling yourself a techno-optimist.” Let’s gooooo!

Andrew Webster
Andrew Webster
Xslop?

The big shakeup at Xbox means that Asha Sharma is taking over for Phil Spencer as CEO of Microsoft Gaming. Her previous role at Microsoft was AI-focused, and here’s what she has to say about utilizing AI for games:

As monetization and AI evolve and influence this future, we will not chase short-term efficiency or flood our ecosystem with soulless AI slop. Games are and always will be art, crafted by humans, and created with the most innovative technology provided by us.

Dominic Preston
Dominic Preston
Samsung’s new Bixby hits beta.

The “conversational” assistant was announced for phones last month, after a rollout on TVs, and is now available in the One UI 8.5 beta in the US and a few other countries. Expect to hear more about it, and its full release plans, at next week’s Unpacked.

Jay Peters
Jay Peters
Nvidia and OpenAI might be getting close to a new, major investment deal.

After a previously-announced $100 billion deal went “on ice,” as The Wall Street Journal reported, Nvidia is nearing a $30 billion equity investment as part of a larger funding round, the Financial Times reports. The investment might be tied up as soon as this weekend.

Jay Peters
Jay Peters
ChatGPT ads in the wild.

Adthena, an “AI search intelligence” platform, has spotted ads in ChatGPT, and they can apparently trigger as soon as after your first prompt.

Ads from Expedia, Qualcomm, Best Buy, and Enterprise Mobility are starting to show up in ChatGPT responses, OpenAI told Adweek.

This is a screenshot from a whole slideshow about the ChatGPT ads on Adthena’s LinkedIn page.
This is a screenshot from a whole slideshow about the ChatGPT ads on Adthena’s LinkedIn page.
Image: Adthena
The Pitt has a sharp take on AI

HBO’s medical drama has been teasing out a smart story about what makes gen AI so tempting and concerning.

Charles Pulliam-Moore
Jay Peters
Jay Peters
The new Gemini 3.1 Pro AI model “represents a step forward in core reasoning.”

Gemini 3.1 Pro is rolling out starting today in the Gemini app and NotebookLM. According to Google:

3.1 Pro is designed for tasks where a simple answer isn’t enough, taking advanced reasoning and making it useful for your hardest challenges. This improved intelligence can help in practical applications — whether you’re looking for a clear, visual explanation of a complex topic, a way to synthesize data into a single view, or bringing a creative project to life.

Jay Peters
Jay Peters
YouTube’s AI chatbot, now on TV.

YouTube is starting to test its conversational AI tool with a “small group of users” on smart TVs, gaming consoles, and streaming devices. The tool, first introduced in 2023, lets you ask questions about the videos you’re watching.

Money no longer matters to AI’s top talent
Play

The AI industry is rife with defections, FOMO, and radical mission statements. It’s about to get supercharged.

Nilay Patel
It’s MAGA v Broligarch in the battle over prediction markets

Prediction: This is going to be a mess for the Trump right.

Tina Nguyen
Dominic Preston
Dominic Preston
The best hot dog eating tech journalists.

Chatbots could bring back the golden age of search spam, as one BBC journalist found after publishing a blog post about his (fake) hot dog eating exploits, which fed right into AI answers. SEO spam is nothing new, but Gemini’s confident attribution to hot dog “reporting” makes things a whole lot worse.

Emma Roth
Emma Roth
OpenAI hires away Instagram’s VP of global partnerships.

Charles Porch, who helped land Instagram’s biggest partnerships — including the launch of Beyoncé’s self-titled album on the social network — will now serve as OpenAI’s first VP of global creative partnerships.

“I’m going to be the person that’s talking to creative communities around the world to figure out how we build the best products to serve them,” Porch tells Vanity Fair.

Stevie Bonifield
Stevie Bonifield
Meta plans to spend $65 million on elections to sway AI legislation.

The funding will go toward Meta’s pro-AI super PACs, including two new ones: Republican-focused “Forge the Future Project” and Democrat-focused “Making Our Tomorrow,” the New York Times reports. The PACs will back politicians who are friendly to AI and push back against legislation that could limit the growth of Meta’s AI business.

Jay Peters
Jay Peters
Epic is acquiring a company that makes AI tech for animating digital humans.

Meshcapade’s team will join Epic Games’ AI Research team and contribute to technologies for Unreal Engine and the hyperrealistic MetaHumans.

With the acquisition, Epic says it’s “looking forward to working together to advance digital human technologies for use across gaming, film and entertainment.”

Update: Added Epic statement.

Max Planck spin-off draws Epic Games to Tübingen

[The Max Planck Society for the Advancement of Science]

Nilay Patel
Nilay Patel
What is Ring’s Search Party feature really for?

A new report from 404 Media today featured a leaked email from Ring founder Jamie Siminoff, who leads the camera maker inside Amazon, saying back in October that he has grander ambitions for the company’s controversial Search Party feature beyond just finding lost dogs.

We had Siminoff on Decoder a few months ago, when I asked him explicitly about using facial recognition to identify people, something the company has since claimed it has no plans to do. Check out what he had to say in the clip below.

Thomas Ricker
Thomas Ricker
Netflix gives ByteDance three days to stop Seedance AI theft.

Otherwise, the TikTok parent will face “immediate litigation” for copyright infringement of Netflix’s Stranger Things, KPop Demon Hunters, Squid Game, and Bridgerton franchises:

“Seedance acts as a high-speed piracy engine, generating mass quantities of unauthorized derivative works utilizing Netflix’s iconic characters, worlds, and scripted narratives. Netflix will not stand by and watch ByteDance treat our valued IP as free, public domain clip art.”