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.
RSL Media — the nonprofit founded by Blanchett — is opening up access to a new registry that will allow people to set terms for how AI systems use their likenesses.
Through the registry, people can permit, prohibit, or require payment for AI systems to use their name, image, voice, and other attributes. It will soon let people signal how AI can use their creative works, characters, and marks, as well.
8
Verge Score
Google Health remains a work in progress, but it could be so much worse.
Anthropic’s most recent update on the outage is from about 10:25AM ET, saying “The issue has been identified and a fix is being implemented,” but that’s all we have for now, other than a dropping number of reports noted by Down Detector.
Of course, its own status dashboard shows this is a pretty common occurrence.
A new “Fix” button gives Workspace users an option to have Gemini to diagnose and automatically resolve formula errors in Sheets.
As Android Police highlights it will pop up any time you click on a cell with an error in it, and show Gemini’s fix in the chatbot’s sidebar.
They’re $80 cheaper, and Alex Himel, Meta’s VP of wearables, says privacy improvements are on the way.
Courtroom Chaos — Starring Arnold Schwarzenegger follows the Snoop Dogg-themed courtroom game that launched last year on Luna, and they’re both part of Amazon’s new plan for games.
This new game adds a single-player mode and a mode where AI Arnold can help you settle bets with friends.
Meta said it didn’t have indicators of the post being nonconsensual, but in addition to arguing that was incorrect, the board says Meta should determine that “AI-generated impersonation is non-consensual by default.”
Screenshots seen by Business Insider showed that data from Meta’s controversial AI training program — including employees’ private conversations, performance data, and transcriptions — could be accessed across the entire company. In a statement to the publication, Meta said:
“We have carefully designed this program with privacy safeguards, and while we have no indication at this time that any data was improperly accessed by Meta employees, we’re pausing it while we investigate.”
Yup, that about sums up where we are right now in the world of AI editing tools and content authenticity.
That’s a workforce reduction of about 13 percent over the last 12 months, according to a disclosure filed Monday. And Oracle is hardly alone in at least partially attributing the cuts to AI advancements:
Worries are quickly mounting over job losses due to AI disruption, as 196 tech companies laid off more than 119,800 employees so far this year, according to Layoffs.fyi.
Japanese AI startup Sakana has posted benchmark results and beta tester impressions saying that its Fugu models compare favorably with or beat the likes of Fable 5, Gemini 3.1 Pro, and GPT-5.5.
The company says Fugu carefully chooses when to use other frontier AI models like Claude and Gemini on specific tasks, although it doesn’t tell users which model was used for what. “Over time, Sakana Fugu will naturally grow by incorporating newer, more efficient models, including our own.”
The open-source AI startup has a deal through 2029, similar to Anthropic and Google, to rent compute capacity from SpaceX’s Colossus 2 data center for $150 million per month (up to $6.3 billion), reports the Wall Street Journal.
[Wall Street Journal]
The intelligence-sharing alliance says that Al models are anticipated to fundamentally transform offensive and defensive cyber capabilities in a matter of months, and that “breaches will occur” as previously unknown vulnerabilities emerge. “Adversaries are already using AI to move faster and more effectively,” said Five Eyes. “Defenders must do the same.”
[National Cyber Security Centre]
Under a multi-year agreement, ChatGPT will surface photos from Getty Images in its responses, as well as within its search results. Getty Images has a similar deal with Perplexity.
Your dream vibe-coded app might be a security nightmare.
Chipmaker SK Hynix, which supplies high-bandwidth memory (HBM) chips to Nvidia and Google, has overtaken Samsung, with Reuters reporting that its market capitalization has now hit $1.35 trillion. Samsung had held the top spot since 2000, prior to the AI boom.
In an interview with The Wall Street Journal, the Microsoft CEO says that people won’t tolerate a small group of companies “doing all of the learning for the world.” Insert the We’re-All-Trying-To-Find-The-Guy-Who-Did-This meme. According to Nadella:
“You can’t say, hey, all white-collar jobs are gone and this could even be a weapon and we will use all the power to build data centers.”
“I hope u have the life u deserve,” is just vicious.
According to Variety, she posted a story dismissing AI-supporting musicians after taking to Instagram earlier to claim that 238 of her songs appeared in AI training data, likely referring to the searchable database posted by Atlantic.
As spotted earlier by The Register, CrankGPT is a concept project from a two-person team at Squeez Labs that runs private, local AI models without needing a power-hungry data center. Instead, you have to literally crank out your own power for each of the chatbot’s answers.
John Jumper, who has worked as a researcher at Google DeepMind since 2017, announced his departure from the company on X. In 2024, Jumper and Google DeepMind CEO Demis Hassabis won the Nobel Prize in Chemistry for developing an open-source AI model that predicts protein structures.
As Reuters reports, the new restrictions will go into effect in August, limiting how students in different age groups use AI:
Pupils from first through seventh grade, aged 6 to 13, should as a general rule not be using AI, while those in lower secondary school, aged 14 to 16, can cautiously adopt tools under teachers’ supervision, the government said. In upper secondary education, from ages 17 to 19, students should learn to use AI appropriately so that they are prepared for further education and work, it added.
“Think of it like a driving instructor with dual controls,” Google’s blog post stated. “The instructor trusts the student but stays ready to take the wheel or hit the brakes if a mistake occurs.” Google DeepMind’s plan itself lays out “internal guardrails designed to catch potential adversarial behaviour by AI agents, even as they become increasingly harder to oversee and contain,” naming methods like chain-of-thought monitoring, asynchronous alerts, real-time access control, and shutdown infrastructure.
[Google DeepMind]
“AIs that pretend to be human are, you know, manipulative,” Jonze said Wednesday at Replit’s Vibecon conference in New York City. “The kids need to grow up knowing these are going to be very, very convincing and very seductive — and very useful and very powerful — but they’re still just a system, an incredible system of pattern recognition.”
He rejoined the company in January after a stint as cofounder of Mira Murati’s competitor, Thinking Machines Lab.
After speaking up for regulation on data centers, Seattle activists say they were called into meetings with HR.
With the Mythos debacle, Anthropic gets its first taste of the Trump admin’s new AI regulation regime.
Noam Shazeer spent twenty years at Google before returning in 2024 after the Big G reportedly paid Character.AI — a company co-founded by Shazeer in 2021 — $2.7 billion to bring him and a team of researchers back home.
The Cybersecurity and Infrastructure Security Agency (CISA) gained access to the limited release cybersecurity-focused model last week, Nextgov/FCW reports. It’s just a little late, since the rest of the world has mostly moved onto the drama around the Trump administration’s block of the safeguarded public version of the model, Fable.
Alongside the announcement of an updated experience for scheduling tasks, OpenAI said that Pulse would be going away “in the next 14 days” and suggested using scheduled tasks for a daily briefing instead.
Pulse was short-lived: the company launched it in September.
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:
The AI design chatbot’s new editor has controls for directly dragging, resizing, and aligning elements, and more options for apps you can export to, including Adobe and Canva. Users can also work on design projects directly from the Claude Code terminal, or hand off software layouts from Design directly to Claude Code, where it picks up exactly where you left off, without a screenshot or a rebuild from scratch.
