Anthropic has introduced Claude Opus 4 and Claude Sonnet 4, its latest generation of hybrid-reasoning AI models optimized for coding tasks and solving complex problems. 

Claude Opus 4 is Anthropic’s most powerful AI model to date, according to the company’s announcement, and capable of working continuously on long-running tasks for “several hours.” In customer tests, Anthropic said that Opus 4 performed autonomously for seven hours, significantly expanding the possibilities for AI agents. The company also described its new flagship as the “best coding model in the world,” with Anthropic’s benchmarks showing that Opus 4 outperformed Google’s Gemini 2.5 Pro, OpenAI’s o3 reasoning, and GPT-4.1 models in coding tasks and using “tools” like web search.

Claude Sonnet 4 is a more affordable and efficiency-focused model that’s better suited to general tasks, which supersedes the 3.7 Sonnet model released in February. Anthropic says Sonnet 4 delivers “superior coding and reasoning” while providing more precise responses. The company adds that both models are 65 percent less likely to take shortcuts and loopholes to complete tasks compared to 3.7 Sonnet and they’re better at storing key information for long-term tasks when developers provide Claude with local file access.

A new feature introduced for both Claude 4 models is “thinking summaries,” which condenses the chatbots’ reasoning process into easily understandable insights. An “extended thinking” feature is also launching in beta that allows users to switch the models between modes for reasoning or using tools to improve the performance and accuracy of responses.

Claude Opus 4 and Sonnet 4 are available on the Anthropic API, Amazon Bedrock, and Google Cloud’s Vertex AI platform, and both models are included in paid Claude plans alongside the extended thinking beta feature. Free users can only access Claude Sonnet 4 for now.

In addition to the new models, Anthropic’s Claude Code agentic command-line tool is now generally available following its limited preview in February. Anthropic also says it’s shifting to provide “more frequent model updates,” as the company tries to keep up with competition from OpenAI, Google, and Meta.

After directing dozens of documentaries over his 60-year career, legendary filmmaker Martin Scorsese will now have his own life chronicled for Apple TV Plus. In its announcement, Apple says the five-part Mr. Scorsese documentary series will explore how themes like “the place of good and evil in the fundamental nature of humankind” have shaped Scorsese’s filmography as far back as his student work at New York University.

Apple hasn’t mentioned a release date for the docuseries, which is being directed by Rebecca Miller (She Came to Me, Personal Velocity). Mr. Scorsese will benefit from “exclusive, unrestricted access to Scorsese’s private archives,” according to Apple, alongside extensive interviews with Scorsese himself that dive into how his own life experiences have influenced his work. 

The series also includes “never-before-seen interviews” with the filmmaker’s friends, family, and creative collaborators, including Robert De Niro, Leonardo DiCaprio, Mick Jagger, Robbie Robertson, Steven Spielberg, and Miller’s husband, Daniel Day-Lewis.

“[Scorsese’s] work and life are so vast and so compelling that the piece evolved from one to five parts over a five-year period,” said Miller. “Crafting this documentary alongside my longtime collaborators has been one of the defining experiences of my life as a filmmaker.”

The mysterious device that OpenAI is cooking up with former Apple designer Jony Ive will be pocket-size, contextually aware, screen-free, and isn’t eyewear. Ive and OpenAI CEO Sam Altman revealed details about the project in an internal staff call reviewed by The Wall Street Journal, after announcing the $6.5 billion acquisition of Ive’s AI hardware startup, io.

Altman suggested that the acquisition could increase OpenAI’s value by $1 trillion, and envisioned a “family of devices” being born from the partnership. Information about the first device, which Altman is aiming to release by late 2026, has been kept tightly under wraps since its development was confirmed last year over concerns that competitors will set about trying to copy the product before it’s launched to the public. 

Altman dropped some hints during the call that shape our expectations, however, including that it will be unobtrusive, fully aware of a user’s life and surroundings, and will serve as a “third core device” a person would put on a desk after a MacBook Pro and an iPhone. OpenAI is already predicting that the device will be popular, with Altman saying that it will ship “faster than any company has ever shipped 100 million of something new before.”

The leaked call also gave some insight into what the device likely won’t be — Altman said that it isn’t a pair of glasses, and that Ive wasn’t keen to make something you’d need to wear on the body, having recently slammed the Humane AI Pin. Altman has also denied rumors that OpenAI is developing a phone. The Journal previously reported that Ive and Altman wanted to wean users away from screens, with Ive saying in a recent interview that his next product is driven by owning the “unintended consequences” associated with the iPhone.

Altman told OpenAI employees on the call that they have “the chance to do the biggest thing we’ve ever done as a company here.” The Journal reports that Ive referred to the project as “a new design movement,” and harkened back to his Apple career that saw him work closely with Steve Jobs before his passing in 2011. Now teamed up with Altman, Ive said, “the way that we clicked, and the way that we’ve been able to work together, has been profound for me.”

Google didn’t want to give publishers the choice to keep their content out of AI Search results because it’s “evolving into a space for monetisation.” That’s according to a newly disclosed internal document, spotted by Bloomberg, which reveals that Google had discussed offering publishers more granular control over how website data would be used in AI Search features instead of the illusion of choice they eventually received.

The document, written by Google Search executive Chetna Bindra, was released during the US antitrust trial into Google’s online search monopoly. The access to its search engine data gives Google a huge advantage in AI development over rivals like Perplexity and OpenAI. But Google’s AI Overviews and AI Mode can be detrimental to the websites they source from by reducing clickthroughs, incentivizing publishers to keep their content out of AI summaries and related features if given the choice.

One of the suggestions in the documents that Google considered a “hard red line” would enable publishers to prevent Google’s AI models from referencing their data in real time, but not opt out of being used to train features like AI Overviews generally. Another option, labeled as “likely unstable,” suggested that no additional controls should be added, and that publishers can opt out of being indexed on Search entirely “if not satisfied.”

A court hearing on May 2nd revealed that publishers are facing that ultimatum. While Google introduced a way for publishers to opt out of AI training in 2023, Google DeepMind vice president of product Eli Collins said it doesn’t apply to search-specific AI products like AI Overviews. The only way for publishers to avoid AI Overviews sucking up their content is to opt out of being crawled by Googlebot — which stops their website being indexed for Search altogether.

When AI Overviews rolled out last year, Google decided to “silently update” the information about publisher controls with “no public announcement,” according to the document. Guidance on how to word the update also suggests that Google intentionally made it harder for publishers to know what they were actually opting out of to avoid getting “into the details of distinction” between training for Gemini, AI Overviews, and other AI models.

“Do what we say, say what we do, but carefully,” Bindra said in the document.

Google says that this document was an early list of options it was considering as AI search was evolving, and doesn’t reflect the decisions it ultimately made. “Publishers have always controlled how their content is made available to Google as AI models have been built into Search for many years, helping surface relevant sites and driving traffic to them,” Google spokesperson Peter Schottenfels said in a statement to The Verge. “New search features like AI Overviews have led to more searches, which creates new opportunities for sites to be discovered.”

The wording that Google currently uses is more upfront, saying that publishers who flag their content not to be used for AI Overviews and AI Mode will also keep it out of “all forms of search results.”

Strava is making it easier to plan workouts and keep activity leaderboard rankings fair. The updates rolling out over the coming weeks focus on helping users optimize their workout routes to compete against other users and their own personal bests, building on some of the existing AI features that Strava announced last year.

Anyone paying for a Strava subscription (starting at $11.99 monthly) can now access a new AI-powered routes experience under the Maps tab that should provide more intuitive suggestions based on popular routes enjoyed by other Strava users. Users can generate community-backed routes from custom starting points or their current location, pulling data from Strava’s heatmaps feature.

Other route-related updates will be rolled out to the Strava mobile app in the coming months, including changes to the tappable points of interest (POI) feature that currently enables subscribers to instantly generate routes to cafés, restrooms, viewpoints, and other locations. Starting in June, POI’s will also display elevation, distance, and estimated arrival time information, and allow users to upload photos of the location. Point-to-point routing will also launch in July, which uses heatmaps and machine learning to deliver “the most efficient, activity-specific route from A to B,” according to Strava.

Strava is also doubling the number of live segments, which allow users to view real-time performance data and achievements in sections of their route, and introducing additional data screens for subscribers.

Finally, Strava says it’s “continuing to advance” the AI-powered Leaderboard Integrity feature it launched to weed out cheaters on cycling and running paths. The company says that 4.45 million activity logs have been removed so far that carried the wrong sport type, or were recorded in vehicles — which is an easy way to fabricate scores now that e-bikes can make anyone the king of a mountain.

Ad-supported tiers are proving to be popular with streaming customers. New data from subscription analyst firm Antenna shows that 46 percent of Discovery Plus, Disney Plus, Hulu, HBO Max, Netflix, Paramount Plus, and Peacock subscribers in the US are paying for ad-supported plans, and that around 75 percent of subscribers have at least tried them.

Many streaming services have ad-supported tiers  — HBO Max launched its $9.99 ad plan in 2021, followed by Netflix’s $6.99 and Disney Plus’s $7.99 plans in 2022, for example. But in its Q2 2025 State of Subscriptions report, Antenna notes that half of the big streaming platforms it analyzed didn’t offer an ad-supported plan two years ago, and only a third of subscriptions to services that did were for an ad plan. Comparatively, 71 percent of net subscriber additions over the last nine quarters have been driven by ad plans according to Antenna’s data, with no meaningful differences in demographic and loyalty compared to ad-free subscribers.

It’s a win-win for streaming companies that can reap the benefits of both additional advertising revenue and growth from providing more affordable memberships. Antenna reports that 65 percent of users who had subscribed to ad-supported plans were completely new to the streaming service, with users who had switched from pricer ad-free tiers accounting for only 11 percent of subscriptions.

Netflix’s ad tier has especially taken off, having doubled in subscribers over the last year. That success is emboldening the platform to start experimenting with its advertising model, with interactive and pause screen ads coming in 2026, and plans to blend AI ads into shows and movies in the future.

Some of Adobe’s most expensive Creative Cloud subscriptions are about to get even pricier for users in North America. Starting from June 17th, the Creative Cloud All Apps plan will be renamed Creative Cloud Pro for users in the US, Canada, and Mexico, adding a bunch of generative AI perks in exchange for bumping up subscription costs.

The pre-tax monthly price for individual Creative Cloud All Apps subscribers on an annual contract will increase from $59.99 to $69.99, or from $659.88 to $779.99 annually. The monthly price for rolling, non-contracted subscribers will jump from $89.99 to $104.99. Contracted prices for teams start at $99.99 per month, up from $89.99, while student and teacher plans will jump from $34.99 to $39.99 monthly on renewal.

The new Creative Cloud Pro plan provides access to the same Adobe apps and features as the previous All Apps plan, alongside some new additions. Users will have unlimited credits for powering generative AI image tools like Photoshop’s Generate Fill, and 4,000 monthly credits for “premium” AI video and audio features like Generative Extend in Premiere Pro. The plan also includes the ability to select third-party generative AI models like OpenAI’s GPT and Google Imagen, and access to Adobe’s in-beta collaborative whiteboard app, Firefly Boards.

“We want to empower you with more time to explore ideas and create, so we’re focused on continually upgrading your apps’ performance, delivering innovations to your core workflows, and integrating generative AI-powered capabilities across Creative Cloud and through our new Firefly app that make you more productive,” Adobe said in its announcement. “Today, we’re sharing updates to our Creative Cloud offerings to reflect our continued innovation and commitment to providing you the best tools for bringing your visions to life.”

The previous All Apps plan will no longer be available following the switch, and existing subscribers will be charged the increased price on their next renewal date. These changes only apply in North America, and Adobe says it isn’t planning to make name or pricing changes in other regions “at this time.”

Adobe is also offering a stripped-back “Creative Cloud Standard” plan for users who don’t need its generative AI offerings. Starting at $54.99 per month for contracted subscribers or $82.49 for rolling users, the Creative Cloud Standard plan is nearly identical to the current All Apps plan, but reduces the previous 1,000 monthly generative credit allowance down to 25. Users also won’t have access to premium web and mobile app features that are unlocked for Creative Cloud Pro subscribers. 

Not only is Adobe defaulting current All Apps subscribers to the more expensive AI-laden plan, compared to the Standard offering that’s closer in price and features, but this new Creative Cloud Standard offering is only available to existing subscribers, forcing new customers to take the more expensive AI-focused plan.

Adobe’s All Apps plan was a poor investment for most individual users anyway, as it’s rare for one person to need more than 20 apps that target such a wide variety of creative industries. Still, some existing All Apps subscribers aren’t too pleased with the changes, if the reactions on Adobe subreddits are any indication. Canva attempted to make similar AI-driven price increases last year, which were later softened due to backlash from its users, so we will have to see if Adobe sticks to its guns.

Elgato announced a new “Stream Deck Everywhere” strategy at Computex, which includes a software-based virtual Stream Deck for PCs, an ethernet dock accessory, unbranded Stream Deck modules that can bring its keys into other products, and even new scissor-switch keys for a more precise, keyboard-like feel. “With these launches, we’re opening up Stream Deck to a world of new users, from developers and DIY builders to manufacturers seeking a field-proven interface for their products,” said Elgato general manager Julian Fest. “Because Stream Deck isn’t just a device—it’s a platform.

Stream Deck Modules are designed for hobbyists and manufacturers to easily integrate into hardware projects instead of ripping apart the consumer version or developing their own custom macropad solutions. The modules are available in 6-, 15-, and 32-key variants and are housed in an aluminum chassis that can be built into custom bases, machines, and furniture, making it possible to actually create Elgato’s ridiculous 1,262-key April Fools’ desk.

The Stream Deck Modules are available now starting at $49.99 for the six-key version, $129.99 for the 15-key, and $199.99 for the 32-key. Elgato technical marketing manager Philipp Eggebrecht told The Verge that discounts will be offered for bulk purchases.

A new variant of the MK.2 Stream Deck has been introduced that replaces the membrane keys with scissor-style switches that provide “enhanced speed and precision,” according to Elgato. Eggebrecht says the scissor keys variant costs $149.99 and will be available in “around two weeks.” 

Stream Deck buttons that feel more like keyboard keys would make it easier for users to tell they’ve clicked something without having to look at the device. Elgato describes the scissor switch variant as something that “lends itself to typing-style interactions, paving the way for new use cases that benefit from rapid, multi-layered input and control,” which could be testing the waters for a Stream Deck keyboard or any other application requiring precise controls.

“The idea of a full LCD-key based keyboard is interesting. We’ve explored it in the past and so far, what we’ve found is that the experience hasn’t been great, and there’s the cost question,” Eggebrecht told The Verge. “We’re getting closer to a keyboard feeling with Stream Deck Scissor Keys, and always looking at the technology and what is becoming possible. Stream Deck and keyboards are a natural fit.”

Elgato is also rolling out a software-only Stream Deck for desktop devices, akin to the existing Stream Deck mobile app. The Virtual Stream Deck (VSD) can provide a permanent macro menu on computer screens or be summoned at will when users hit allocated hotkeys and mouse buttons. The VSD software will initially be available for people who own a Corsair Scimitar Elite Wireless SE mouse, Xenon Edge touchscreen, or any Stream Deck model, before later rolling out to “more devices.”

Custom layouts, virtual faceplates, and support for beyond six buttons require iPhone and Android Stream Deck users to pay for Pro access, which starts at $2.99 monthly or $49.99 as a one-time purchase. Eggebrecht says Elgato’s desktop VSD isn’t locked behind a subscription, but does require users to install the Stream Deck 7.0 beta.

A new $79.99 Stream Deck network dock accessory was also announced that includes support for power over Ethernet (PoE), which lets you place its buttons anywhere there’s a network jack, and with more flexibility than USB-C affords by itself. Eggebrecht says the Stream Deck network dock will start shipping in August.

23andMe will keep offering customers its DNA testing services after being bought out of bankruptcy. New York-based biotech company Regeneron Pharmaceuticals announced an agreement on Monday to purchase the 23andMe startup for $256 million, alongside its Total Health and Research Services business and biobank of customer data and genetic samples.

Regeneron is the winner of 23andMe’s bankruptcy auction, which required all bidders to comply with applicable laws and the firm’s privacy policies around customer data. 23andMe says that customer data is anonymized and that stored genetic samples are destroyed when users delete their 23andMe accounts, but it’s unclear how much information is retained and may, therefore, soon be in Regeneron’s hands.

The acquisition is expected to close later this year, subject to US Bankruptcy Court approval. If all goes ahead, Regeneron co-founder George D. Yancopoulos says the purchase will further the company’s “large-scale genetics research” into future drugs and treatments.

23andMe has collected genetic samples and data from more than 15 million customers since launching its at-home DNA testing kit business. Once briefly valued at $6 billion after going public in 2021, the company filed for bankruptcy in March after failing to turn a profit. Its co-founder and former CEO, Anne Wojcicki, simultaneously stepped down from the company.

“We are pleased to have reached a transaction that maximizes the value of the business and enables the mission of 23andMe to live on, while maintaining critical protections around customer privacy, choice, and consent with respect to their genetic data,” said 23andMe chair Mark Jensen. “We are grateful to Regeneron for offering employment to all employees of the acquired business units, which will allow us to continue our mission of helping people access, understand, and gain health benefits through greater understanding of the human genome.”

Anthropic has responded to allegations that it used an AI-fabricated source in its legal battle against music publishers, saying its Claude chatbot made an “honest citation mistake.”

An erroneous citation was included in a filing submitted by Anthropic data scientist Olivia Chen on April 30th, as part of the AI company’s defense against claims that copyrighted lyrics were used to train Claude. An attorney representing Universal Music Group, ABKCO, and Concord said in a hearing that sources referenced in Chen’s filing were a “complete fabrication,” and implied they were hallucinated by Anthropic’s AI tool.

In a response filed on Thursday, Anthropic defense attorney Ivana Dukanovic said that the scrutinized source was genuine and that Claude had indeed been used to format legal citations in the document. While incorrect volume and page numbers generated by the chatbot were caught and corrected by a “manual citation check,” Anthropic admits that wording errors had gone undetected.

Dukanovic said, “unfortunately, although providing the correct publication title, publication year, and link to the provided source, the returned citation included an inaccurate title and incorrect authors,” and that the error wasn’t a “fabrication of authority.” The company apologized for the inaccuracy and confusion caused by the citation error, calling it “an embarrassing and unintentional mistake.”

This is one of many growing examples of how using AI tools for legal citations has caused issues in courtrooms. Last week, a California Judge chastised two law firms for failing to disclose that AI was used to create a supplemental brief rife with “bogus” materials that “didn’t exist.” A misinformation expert admitted in December that ChatGPT had hallucinated citations in a legal filing he’d submitted.

Pinterest has apologized for a recent wave of “over-enforcement” that erroneously deactivated many accounts. The platform has experienced some weird moderation issues in recent weeks, and outraged users reported their accounts had been suspended without warning or explanation. In response to many appeals, the platform cited unspecified community guideline violations.

The company initially addressed ban concerns with a statement saying that it will “continuously monitor for content that violates our Community Guidelines and accounts with violative content may be deactivated as a result.” This answer did little to soothe outrage from users who were calling for Pinterest to clarify how accounts had violated the platform’s guidelines, and complained that appeals to reinstate mistakenly banned accounts were not being processed, or being rejected without explanation.

On Wednesday, Pinterest issued an updated statement by responding to a support post it had published to X in July last year. “We recently took action on violations of our content policies, but an internal error led to over-enforcement and some accounts were mistakenly deactivated,” Pinterest said on Wednesday in an updated statement attached to an old X post. “We’re sorry for the frustration this caused. We’ve reinstated many impacted accounts and are making improvements to respond faster when mistakes happen going forward.”

Pinterest hasn’t given any specific details about what the “internal error” was, what caused it, or if it has been resolved. Some users reported that Pinterest was also deleting pins for seemingly random and inaccurate content violations, such as images of everyday objects being flagged for “adult content,” leading to speculation that pins and accounts were being reported by an inadequately implemented AI moderation system. Pinterest has since told TechCrunch that AI moderation was not responsible for the error.

Users who were mistakenly suspended in the past few weeks are starting to regain access to their accounts, according to reports on the Pinterest subreddit. Given how clumsily the company has handled the situation, however, some scorned users are in no rush to forgive the platform.

Threads is making it easier for creators to share personal links with their communities and followers. The link-sharing update announced today now allows Threads users to add up to five links to their account bios and provides performance insight that shows how many people have visited links shared in posts and profiles.

Meta previously only allowed a single link to be added to profile bios on Threads, and this update brings the platform in line with Instagram’s link-sharing capabilities. While creators could use services like Linktree, which provides a landing page where multiple links can be displayed, for that single link slot, having the URLs for up to five links directly displayed on profiles will make it easier to see which link options are available, such as newsletters, podcasts, articles, and merchandise shopfronts.

Threads is also working on providing users with new audience management features that compare week-over-week performance data for the number of posts made, total views, replies, and new followers. Meta says that these insight recaps are “coming soon,” alongside personalized tips that will be shown to creators to “inspire new ways to engage your community.”

The latest versions of OpenAI’s multimodal GPT AI models are now rolling out to ChatGPT. OpenAI announced on Wednesday that GPT-4.1 will be available across all paid ChatGPT account tiers and can now be accessed by Plus, Pro, or Team users under the model picker dropdown menu. Free users are excluded from the rollout, but OpenAI says that Enterprise and Edu users will get access “in the coming weeks.”

GPT-4o mini, the more affordable and efficiency-focused model that OpenAI launched last year, is also being replaced with GPT-4.1 mini as the default option in ChatGPT for all users, including free accounts. Users on paid accounts will also see GPT-4o mini replaced by GPT-4.1 mini in the model picker options. 

Both GPT-4.1 and GPT‑4.1 mini are optimized for coding tasks and instruction following, and outperform GPT‑4o and GPT‑4o mini “across the board,” according to OpenAI. Both of the new models support a one million context token window — the amount of text, images, or videos in a prompt that an AI model can process — that greatly surpasses GPT-4o’s 128,000-token limit. OpenAI says that speed improvements also make GPT-4.1 more appealing for everyday coding tasks compared to the powerful OpenAI o3 & o4-mini reasoning models it introduced in April.

GPT-4.1 and GPT-4.1 mini were made available to developers last month, alongside the GPT-4.1 nano model, which OpenAI described as its “smallest, fastest, and cheapest” offering to date. GPT-4.1 Nano is absent from this model rollout announcement, and it’s unclear when — or if — it’ll be generally available in ChatGPT.

Apple is trying to dissuade Europeans from using iOS apps that support alternative payment options by making them look scary. Daring Fireball’s John Gruber spotted that a red exclamation mark icon is being prominently displayed on the App Store listing for Instacar, alongside a message warning users that it doesn’t use Apple’s “private and secure payment system.”

Instacar, a three-year-old Hungarian app for checking the mileage and value of used cars, isn’t available on the US App Store. But in Hungary, in the EU, it’s listed as one of the top five apps in the Business category, with thousands of positive reviews — not a typical service that would warrant heightened caution. The only gripe that Apple seemingly has is that Instacar uses an external payment system, and that App Store payment features like purchase history, Family Sharing, and unified subscription management won’t be available on such purchases.

“When you create an account on a developer’s external website, you may have to provide personal information, including payment information, directly to the developer or third-party partners,” Apple says in a support page linked from its warning message. “You will be trusting the developer, as well as any partners and payment providers they work with, to handle your information based on their privacy and security controls.”

Apple is already facing heat in the EU for its practices around scare tactics. In April, the European Commission issued preliminary findings that Apple “makes it overly burdensome and confusing” for users to install alternative app marketplaces, a process that involves clicking through several scare sheets that ask users to confirm if they want to proceed. 

This App Store warning message has also appeared in the wake of the recent Epic vs Apple ruling that bans Apple from restricting how developers can link to alternative purchase systems. A notable requirement implemented to address Apple’s scare tactics was that the company cannot interfere with consumers choosing to leave an app with anything beyond “a neutral message” about being directed to a third-party site, though that injunction doesn’t apply outside of the US.

TikTok is introducing new accessibility features that make it easier for people with visual impairments to see text and use screen readers on the platform. After adding alt text support for still images in April, TikTok is now testing a feature that automatically adds an AI-generated alt text description to photographs when a creator hasn’t manually added it.

Creators will still have the option to add their own alt text descriptions to images as they’re uploaded or after publishing, providing a means to correct any information in the AI-generated versions. Even if the AI alt text is imperfect, however, users who rely on screen readers to process visual content may find that having it widely applied is preferable to images not having any alt text at all.

TikTok is also launching a new color contrast toggle that can be used to increase the foreground color of text, icons, and UI elements. This should make descriptions and controls more distinct, helping people with low vision or contrast sensitivities to navigate the platform. Additionally, TikTok will now automatically display all text in bold across the app when it detects that users have enabled bold text support on their devices.

All of these features are available now and can be found on TikTok’s redesigned accessibility settings page, which is located on iOS and Android user profiles under the three-line menu, or the left-hand taskbar on desktop.

Nvidia’s efforts to suck up to the Trump administration have seemingly paid off, with the US now lifting export limits on US-made AI chips and cracking down on anyone using Huawei’s emerging alternatives. The announcements come as Nvidia CEO Jensen Huang joined President Trump in Saudi Arabia this week to solicit AI investments for US companies.

The US Department of Commerce (DOC) announced on Monday that it has rescinded the Artificial Intelligence Diffusion Rule, due to take effect on May 15th, that aimed to restrict how many US-made AI chips could be sent to international markets without special government approval. The DOC said that a replacement rule for protecting US AI technology will be issued “in the future,” but provided no specific details.

“These new requirements would have stifled American innovation and saddled companies with burdensome new regulatory requirements,” The DOC said in a statement. “The AI Diffusion Rule also would have undermined US diplomatic relations with dozens of countries by downgrading them to second-tier status.”

While the goal was to prevent countries already subject to chip restrictions, such as Russia and China, from accessing or building AI tech, it also placed Nvidia’s estimated 90 percent share of the AI chip market in jeopardy. Shortly after the Diffusion Rule was introduced by former President Joe Biden in January, Nvidia issued a statement calling it “misguided,” while anticipating a return to Trump’s first term policies “that strengthen American leadership, bolster our economy and preserve our competitive edge in AI and beyond.”

The DOC also warned companies that using Huawei’s Ascend AI chipset “anywhere in the world” would violate US export control agreements. Huawei’s home-grown Ascend processors are seen as China’s best answer to Nvidia’s powerful AI chips. 

Nvidia’s Huang was notably one of the only US tech leaders to not attend Donald Trump’s inauguration. His absence doesn’t appear to have soured the relationship between them, however, with Huang spotted cosying up to Trump at a US-Saudi investment summit in Riyadh on Tuesday, alongside other tech leaders like Elon Musk, AMD’s Lisa Su, OpenAI’s Sam Altman, and Epic CEO Tim Sweeney.

The Washington Post reports that Trump was far from subtle about what the US wanted from the gathering. “As you know, we have the biggest business leaders in the world here,” he told  Saudi Crown Prince Mohammed bin Salman. “They’re going to walk away with a lot of checks for a lot of things that you’re going to provide.”

It would seem that the money-making days for CryptoPunks, one of the oldest and most valuable NFT brands, are well and truly over. After acquiring the non-fungible token collection from Larva Labs in 2022, Bored Ape Yacht Club creator Yuga Labs is now selling CryptoPunks to Infinite Node Foundation — a non-profit focused on preserving digital art.

CryptoPunks, alongside brands like Bored Ape Yacht Club and CryptoKitties, is frequently credited for kicking off the NFT craze in 2021. Several tokens have been sold for millions, including one in 2022 that was purchased for more than $23 million worth of Ethereum cryptocurrency, but the sales charts have largely stagnated since. 

Details for the purchase are slim. CryptoPunks general manager Natalie Stone told The Verge that the terms of the deal won’t be disclosed, and that the brand will be stewarded by the Node Foundation, with no plans to integrate the CryptoPunks token collection into the gamified Otherside metaverse that Yuga Labs is developing. “Yuga is still one of the biggest holders of Punks, and like all other CryptoPunks NFT holders, will retain the right to its CryptoPunks characters under the IP license,” said Stone.

It was also never clear how much Yuga Labs spent to acquire CryptoPunks in its heyday, but being sold to a non-profit is probably not a good sign for, well, profitability. Now that NFT’s are a shambling shadow of their former glory, the focus is turning to making CryptoPunks relevant in the mainstream art industry.

“We’ve worked to elevate and protect their legacy, but we always knew Punks needed a permanent home built for preservation,” Yuga Labs co-founder Wylie Aronow said in the announcement release. “Seeing that vision come to life with the Node Foundation feels like a full-circle moment. They are best positioned to protect the Punks legacy.”

Node Foundation chair Micky Malka said that “CryptoPunks sparked a cultural movement,” and that the non-profit intends to “future-proof this landmark work” and make it more accessible for scholars, curators, and collectors to engage with. Node is planning to showcase all 10,000 CryptoPunks tokens at the foundation’s permanent exhibition space in Palo Alto, and in the “Infinite Images: The Art of Algorithms” exhibit at the Toledo Museum of Art.

The Square Handheld is available today for $399 with support for traditional card transactions and tap-to-pay, scanning products, and managing inventory, all from a single device. That should bring tap-to-pay acceptance without an iPhone to more places in the businesses and restaurants where you’re already used to seeing the Square Stand and other similar setups.

Weighing in at 11 ounces (about 311 grams) and measuring less than an inch thick, Square says this is “one of the thinnest and lightest handheld point of sale devices on the market.” That makes the Handheld easier to lug around than Square’s wedge-shaped Terminal POS device, and provides some competition for Clover’s $749 Flex system. 

The Square Handheld includes a splash and dust-resistant 6.2-inch touchscreen, a barcode scanner, and a 16 megapixel camera. Square says that the battery provides enough juice to “power through a whole day.” Belkin is also offering SheerForce protective cases for the Handheld starting at $39, which can help businesses to better tailor the device to their brand identity and bolster it against damage.

Square has also redesigned its POS app, giving the Handheld access to seven modes that can be customized with industry-specific features such as Quick Service mode, Bar mode, and Retail mode.

Spotify is giving users more control over which songs are spat out by its AI DJ feature. Starting today, Spotify Premium subscribers can now use English voice commands to personalize what music they want DJ to play instead of relying on the ever-changing playlist that it curates based on their listening habits.

Users can access these new capabilities by pressing and holding the ‘DJ’ button located in the bottom right corner when using the AI DJ feature and listening for a beep. Users can then verbally tell the DJ bot what they want to hear, such as “play me some electronic beats for a midday run,” or “surprise me with some indie tracks I’ve never heard before.” 

Prior to the update, this button would just randomly switch up whatever song you were currently listening to and didn’t allow users to control what would be played next. Now, Spotify says that users can direct it to play specific artists, genres, and moods. It also borrows some of the quirkier vibe-based music recommendations from the AI Playlist building feature that Spotify launched in beta last year, which creates a tracklist based on text prompts. For example, users can ask the AI DJ to “play me some music to soundtrack my life as a movie.”

This new voice command update is currently the only means for users to have any control over what tracks the AI DJ will play, however, which is a bit awkward if you’re using the feature in an environment where you can’t be chatty.

Some Google Search users are starting to see the platform’s AI Mode search engine chatbot in the wild. Several user reports across X and Threads show that Google is testing different locations to display the AI Mode tab — while some examples show it in the search bar itself, besides the “search by image” button, in others it replaces the “I’m Feeling Lucky” button underneath the search bar.

There’s also some variation to the appearance of the AI Mode button, with some users seeing a rotating rainbow border when the cursor hovers over it. In another example where AI Mode has replaced “I’m Feeling Lucky,” the rainbow border is seemingly the default design, and makes the search-focused AI chatbot button stand out among Google’s other offerings.

Google said on May 1st that a “small percentage” of people in the US would soon start seeing an AI Mode option in Google Search, but the screenshots of these tests give us some idea of where it may be placed when (or if) it rolls out to the wider public. For now, Google spokesperson Ashley Thompson told The Verge that AI Mode in Search is currently limited to users of Google’s experimental Labs environment.

“We often test different ways for people to access our helpful features,” said Thompson. “This is just one of many experiments.”

It’s not guaranteed that the “I’m Feeling Lucky” option will officially be replaced by AI Mode, but even the knowledge that Google is considering it may upset folks who use the niche feature. The button that takes users directly to the first webpage of any search results has been a fixture of Google’s homepage since the search engine launched. This may be Google’s way of suggesting that AI Mode’s chatbot-like responses to search queries are more efficient than allowing users to bypass the traditional search results page.