MSI has revealed its latest iteration of its Claw PC gaming handheld — and this time, it’s powered by AMD. The company showed off the Claw A8 BZ2EM at Computex 2025, which comes with an AMD Ryzen Z2 Extreme chip along with up to 24GB of DDR5 memory.
That’s a bit less than the 32GB of memory that came with the Intel-equipped Claw 8 AI Plus released late last year, but it still has an 8-inch full HD display, a 120Hz refresh rate, and a 1TB M.2 SSD. The AMD-powered Claw A8 will also come in two colors: white and lime green.
There’s a new MSI Claw 8 AI Plus “Polar Tempest” edition, too, which features an up to Intel Core Ultra 7 258V processor and a 2TB NVMe SSD. It also has what MSI calls a “glittering” white coating. MSI didn’t reveal a release date or price for either model, but it will likely be in the same ballpark as the standard Intel-powered MSI Claw 8 Plus, which Best Buy lists as costing $999.99.
It’s finally possible to purchase an audiobook from Spotify’s iPhone app with just a few taps. On Monday, Spotify announced that Apple approved an update that allows users in the US to see audiobook pricing within the app and buy individual audiobooks outside the App Store.
The update also lets Spotify Premium subscribers purchase additional audiobook listening hours. This change follows last month’s Epic Games vs. Apple ruling, which upended the iPhone maker’s control over the App Store. Under the ruling, Apple can’t collect fees on purchases made outside the app store, nor can it govern how developers point to external purchases.
Spotify submitted the update last week, but now it’s official. The music streaming service pulled audiobook purchases from its iOS app in 2022 after accusing Apple of “choking competition” with App Store rules that made it more difficult to purchase audiobooks. Spotify also started letting iPhone users purchase subscriptions outside the App Store earlier this month.
The iOS apps for Kindle, Patreon, and Delta’s emulator have also taken advantage of the court ruling, but Epic Games is still fighting to bring Fortnite back to the App Store. “This change lowers the barriers for more users to embrace their first — or tenth — audiobook, while allowing publishers and authors to reach fans and access new audiences seamlessly,” Spotify said in its announcement.
Huawei just launched a super sleek folding laptop that might be as thin as your phone. The MateBook Fold, which consists of a single OLED display, is just 7.3mm (~0.3 inches) thick when unfolded and 14.9mm (~0.6 inches) when closed, as spotted earlier by Android Headlines.
To compare, Lenovo’s ThinkPad X1 Fold measures 8.6mm (0.34 inches) thick unfolded and 17.4mm (0.68 inches) when folded. But unlike Lenovo’s device, the MateBook Fold is only available in China for now, with a price of around $3,300.
The MateBook Fold’s 18-inch display folds at a 90-degree angle to form a 13-inch upper screen, mimicking a traditional laptop with a digital keyboard instead of a physical one. The device weighs just 1.16kg (~2.6lbs), with its tandem OLED offering a 3.3K (3296 x 2472) resolution and a peak brightness of up to 1600 nits. The laptop also comes with up to 32GB of RAM and 2TB of storage.
This also marks the debut of Huawei’s in-house operating system, called HarmonyOS 5, on PC. Huawei first launched HarmonyOS on Android, but it has since brought its operating system to PCs after losing access to Microsoft Windows in March due to US sanctions. In addition to coming with the MateBook Fold, the system is available on the new MateBook Pro as well.
GitHub is launching an AI coding agent that can do things like fix bugs, add features, and improve documentation — all on a developer’s behalf. The agent is embedded directly into GitHub Copilot, and it will start working once a user assigns it a task, according to an announcement at Microsoft Build.
To complete its work, GitHub says the AI coding agent will automatically boot a virtual machine, clone the repository, and analyze the codebase. It also saves its changes as it works, while providing a rundown of its reasoning in session logs. When it’s finished, GitHub says the agent will tag you for review. Developers can then leave comments that the agent will automatically address.
“The agent also incorporates context from related issue or PR (pull request) discussions and follows any custom repository instructions, allowing it to understand both the intent behind the task and the coding standards of the project,” GitHub says. The new coding agent is available to Copilot Enterprise and Copilot Pro Plus users through GitHub’s site, its mobile app, and the GitHub Command Line Interface tool.
Microsoft also announced that it’s open-sourcing GitHub Copilot in Visual Studio Code, which means developers will be able to build upon the tool’s AI capabilities.
Correction, May 19th: An earlier version of the article stated the AI coding agent will be available through Copilot Plus, but the service is actually called Copilot Pro Plus.
It’s a great time to catch up on some big games you might’ve missed out on recently. Epic Games is holding a Mega Sale that includes big discounts on a ton of PC games, including Grand Theft Auto V Enhanced, which is on sale for $14.99 instead of $29.99, and The Last of Us Part 1, which costs $29.99 instead of $59.99.
If you don’t know which games you want to get yet, you have some time to decide. The Mega Sale runs from May 15th to June 12th at 11AM ET. Dead Island 2 and Happy Game are also available for free until May 22nd.
Verizon expects to bring fiber to 1 million homes each year following the acquisition. The deal went through after Verizon “committed to ending DEI-related practices,” according to a statement by FCC Chair Brendan Carr.
The Intercept reports that in a May 15th letter to Carr, Verizon’s chief legal officer, Vandana Venkatesh, outlined what it’s walking away from. Because “Verizon recognizes that some DEI policies and practices could be associated with discrimination,” it will no longer have any HR roles or teams focused on DEI, remove references to the term from employee training materials, as well as goals for diversity in its supplies, representation of women and minorities in its workforce. In the letter, Venkatesh says that now Verizon’s public messaging is going to “remove references to ‘DEI’ or ‘diversity, equity and inclusion.’”
When Verizon’s consumer chief, Sowmyanarayan Sampath, appeared on Decoder last month, we asked him about whether it would fight the FCC imposing regulatory requirements against its diversity initiatives with a decade’s worth of lawsuits, the same way it fought net neutrality. It didn’t.
Through the merger, Verizon will also be able to claw back some of its fiber business after it sold parts of its wireline operations, including Fios fiber internet connections, to Frontier in 2015. Carr said the merger will allow fiber to come to more communities, including rural ones. BEAD, a Biden-era initiative, was supposed to pay fiber providers to bring high-speed internet to rural areas, but a report from The Washington Post suggests that the “money isn’t flowing.”
Update, May 16th: Added additional details from Verizon’s letter to the FCC and Decoder.
A system prompt is a set of instructions served to a chatbot ahead of a user’s messages that developers use to direct its responses. xAI and Anthropic are two of the only major AI companies we checked that have made their system prompts public. In the past, people have used prompt injection attacks to expose system prompts, like instructions Microsoft gave the Bing AI bot (now Copilot) to keep its internal alias “Sydney” a secret, and avoid replying with content that violates copyrights.
In the system prompts for ask Grok — a feature X users can use to tag Grok in posts to ask a question — xAI tells the chatbot how to behave. “You are extremely skeptical,” the instructions say. “You do not blindly defer to mainstream authority or media. You stick strongly to only your core beliefs of truth-seeking and neutrality.” It adds the results in the response “are NOT your beliefs.”
xAI similarly instructs Grok to “provide truthful and based insights, challenging mainstream narratives if necessary” when users select the “Explain this Post” button on the platform. Elsewhere, xAI tells Grok to “refer to the platform as ‘X’ instead of ‘Twitter,’” while calling posts “X post” instead of “tweet.”
Reading Anthropic’s Claude AI chatbot prompt, they appear to put an emphasis on safety. “Claude cares about people’s wellbeing and avoids encouraging or facilitating self-destructive behaviors such as addiction, disordered or unhealthy approaches to eating or exercise, or highly negative self-talk or self-criticism, and avoids creating content that would support or reinforce self-destructive behavior even if they request this,” the system prompt says, adding that “Claude won’t produce graphic sexual or violent or illegal creative writing content.”
The US Army is done relying on contractors to repair its equipment. Earlier this month, Army Secretary Daniel Driscoll committed to including right-to-repair provisions in all existing and future contracts with manufacturers, a change Sen. Elizabeth Warren (D-MA) told The Verge will “put an end to our dependence on giant defense contractors who charge billions of dollars and take months to repair critical equipment.”
For now, only the Army has committed to securing right-to-repair provisions in contracts. But Warren is pushing for other military branches to adopt the requirement, addressing long-standing repairability problems across the armed forces. She’s also hopeful that it could have a broader impact across industries and serve as a model for how other companies and organizations can advocate for similar repair-friendly provisions.
For years, reports have highlighted the US military’s struggle to fix its own equipment, forcing it to wait on defense contractors to service them — even when stationed in foreign countries. A 2019 report from The New York Times described how a maintenance Marine in South Korea couldn’t repair a generator needed for training “because of the warranty,” despite having tools to fix them.
The same report said that engines at a US military base in Okinawa, Japan, “were packed up and shipped back to contractors in the United States for repairs,” while ProPublica found that the Navy’s contract with General Dynamics and Lockheed Martin forced the US military to fly contractors to the ship to make repairs on “proprietary” equipment, “adding millions in travel costs and often delaying missions.”
Warren pushed the Army to take a tougher stance on right-to-repair through 2024, calling out the “costly restrictions” that prevent the military from fixing its own equipment in a letter to the Department of Defense. Warren and Rep. Marie Gluesenkamp Perez (D-WA) later introduced the Servicemember Right-to-Repair Act, which would require contractors to provide the military with “fair and reasonable access” to parts, tools, and information needed to repair equipment. The bill was introduced in December 2024 but has not reached the House floor yet.
The US military is a major force for manufacturers to contend with — and Warren hopes that adopting repairability rules will have a ripple effect far outside it. “The Army’s commitment to right to repair shows other industries that they can do the same,” Warren says. “If it can happen here, it can happen in farm equipment, farmers, washing machines, consumer electronics… and every place else that big manufacturers have tried to take two bites at the apple: the initial price and another bite at the consumer to cover subsequent repairs.”
Charter and Cox have announced plans to merge in a $34.5 billion deal that will create a cable and internet behemoth. The two telecom companies say the merger will allow them to “aggressively compete” against larger broadband companies and mobile providers that have rolled out internet plans of their own.
Charter, which currently has 31.5 million customers, and Cox, which has 6.5 million, both face an increasing threat from streaming services like Netflix. Sports-focused streaming packages like those offered by Comcast, DirecTV, Fox, and soon, ESPN, also let viewers get their sports fix without a cable subscription.
The combined company will change its name to Cox within a year after the deal closes, while Spectrum will become the name of Cox’s consumer-facing brand. As part of the deal, Cox customers will get Charter’s “simple and transparent pricing and packaging structures” with no annual contracts, as well as credits for outages lasting longer than two hours. The companies will continue to offer TV, internet, and mobile services.
Cox and Charter don’t say when they expect the deal to close, but it will require approval from Federal Communications chair Brendan Carr, who has suggested that his agency won’t approve mergers if the companies have policies related to diversity, equity, and inclusivity (DEI).
“This combination will augment our ability to innovate and provide high-quality, competitively priced products, delivered with outstanding customer service, to millions of homes and businesses,” Charter CEO Chris Winfrey said in the press release. “We will continue to deliver high-value products that save American families money, and we’ll onshore jobs from overseas to create new, good-paying careers for U.S. employees.”
TikTok will interrupt teens’ feeds with guided meditation after 10PM. The platform has been testing the feature with a “wind down” prompt, but now it’ll show the notification with guided meditation by default to all users under 18.
TikTok’s wind-down prompt initially only played calming music to remind teens to take a break from the app at nighttime. It was also only shown to teens under 16. But now, the prompts will include meditation, too, and will be available to adults who toggle on “Sleep Hours” from the Screen Time settings page as well.
While testing its meditation feature, TikTok says it found 98 percent of users under 16 kept it on after trying it for the first time. Teens can ignore the initial 10PM reminder, but the app will display another full-screen prompt later that’s harder to dismiss.
TikTok has until June 19th to divest its US operations or face a ban, but that hasn’t stopped it from trying to address safety concerns. In March, TikTok also launched a Time Away tool that parents or caregivers can use to prevent their child from accessing the app at certain times.
Coinbase says cyber criminals “bribed and recruited” support workers to help steal customer data and trick victims into sending money to attackers. As a result of the attack, bad actors obtained the names, addresses, phone numbers, government IDs images, account data, and partial social security numbers of a “small subset of users,” according to a blog post on Thursday.
In a filing with the Securities and Exchange Commission, the crypto exchange said it received an email on May 11th from a threat actor who claimed they had information about certain Coinbase accounts. The bad actor demanded $20 million in exchange for not publicly exposing the information, but Coinbase refused to pay.
Coinbase is working with law enforcement to investigate the incident. It also “immediately terminated the personnel involved.” The company “will press criminal charges.”
The crypto exchange notes that the attackers didn’t get login credentials, 2FA codes, or private keys, and weren’t able to access any Coinbase accounts or wallets. Coinbase says it could spend $180 million to $400 million repaying impacted customers. It’s also offering a $20 million reward to anyone who provides information leading to an arrest.
“Scammers — related to this incident or not — may pose as Coinbase employees and try to pressure you into moving your funds,” the company says in its blog post. “Remember, Coinbase will never ask for your password, 2FA codes, or for you to transfer assets to a specific or new address, account, vault or wallet.”
YouTube now has a chart that tracks the top 100 podcasts in the US. The Joe Rogan Experience took the number one spot during the week of May 5th to the 11th, followed by Kill Tony, Rotten Mango, and 48 Hours.
YouTube will update the chart every Wednesday and rank each show by watchtime, according to a blog post from YouTube’s Steve McLendon. The chart will include playlists “designated as ‘podcasts’ by the creator in the upload process, and will not include playlists that contain only clips or Shorts.”
Over the years, YouTube has established itself as one of the top platforms for listening to — or watching — podcasts. In February, YouTube announced that more than 1 billion people tune into podcasts on the platform each month, far more than the 100 million regular podcast listeners Spotify reported in 2023.
YouTube will continue adding more features to the chart and expand it to more regions in the future, McLendon says.
Google is rolling out a handy update for Chrome on Android, as it will now allow you to zoom in on text without affecting the appearance of the webpage. You can use a slider to enlarge text, and then set it for one page or for all the sites you visit.
Previously, when you zoomed in on a page in Chrome’s Android app, it enlarged the entire page, making the site more difficult to navigate. You can try out the updated zoom feature by tapping the three-dot menu in the top-right corner of Chrome and then selecting how far you want to increase the size of the text.
Additionally, Chrome’s Optical Character Recognition tool, which converts an image of text into a machine-readable format, will now automatically recognize scanned PDFs on desktop, letting you highlight, copy, and search for text, as well as use a screen reader with them. Google first started rolling out this feature in beta earlier this year, but now it’s available to all users on desktop.
Google is also expanding Gemini’s integration with TalkBack on Android, a screenreader that uses AI to identify and describe images on your screen. Now, TalkBack will let you ask follow-up questions about an image, such as what color an object is, what material it’s made of, and what else is in the image.
The company is launching Expressive Captions as well, which Google says “provides real-time captions for anything with sound across most apps on your phone — using AI to not only capture what someone says, but how they say it.” That means you’ll see captions that tell you when someone is saying “noooooo” instead of “no.” Google is adding captions for more kinds of sounds as well, like whistling or someone clearing their throat. This feature is rolling out in English in the US, UK, Canada, and Australia on devices running Android 15 and later.
Netflix is working on a new type of ad that will use AI to let advertisers “marry” their product with the streaming service’s shows and movies. This may sound like Warner Bros. Discovery’s plan to turn its IP into commercials, but during Netflix’s Upfront event on Wednesday, the company demonstrated an example that placed the image of a product over a background inspired by one of its shows, like Stranger Things.
If you’re on Netflix’s ad-supported plan, that means you might see an ad that blends in with the show you’re watching, whether it’s Bridgerton or Wednesday. Netflix says advertisers can insert this new ad format in the middle of what you’re watching, or put them on the screen when you hit pause. They could contain an overlay or call to action when they roll out by the end of this year.
It’s not clear how Netflix plans to evolve these ads down the road, but its advertising president Amy Reinhard said the pace of progress “is going to be even faster.” During the event, Netflix also announced that its ad-supported tier reached 94 million monthly users, more than doubling the 40 million it had at the same time last year.
President Donald Trump told Apple CEO Tim Cook that he doesn’t want the company to move iPhone production to India. “I had a little problem with Tim Cook yesterday,” Trump said during his visit to Qatar. “You’re coming in with $500 billion, but now I hear you’re building all over India… We’re not interested in you building in India.”
Trump claimed Apple is going to be “upping their production” in the US following his chat with Cook. In February, Apple promised to invest $500 billion in the US over the next four years, but iPhone production is unlikely to come stateside due to a lack of labor, suppliers, and manufacturing facilities in the US.
In recent years, Apple has ramped up iPhone production in India in response to tension between the US and China. But following Trump’s tariffs announcement, The Financial Times reported in April that Apple plans to source all iPhones sold in the US from India by the end of 2026.
Apple is officially rolling out the next generation of CarPlay — and it’s called CarPlay Ultra. The update is available with new Aston Martin vehicle orders in the US and Canada, but it will be rolling out to existing models with supported infotainment systems in the “coming weeks.” Carmakers like Hyundai, Kia, and Genesis are also working on adding CarPlay Ultra, according to Apple.
CarPlay Ultra also lets drivers use onscreen controls, physical buttons, or Siri to manage a car’s radio and climate. It also adds widgets “powered by the iPhone that perfectly fit the car’s screen or gauge cluster to provide information at a glance.“ CarPlay Ultra works with the iPhone 12 or newer with iOS 18.5 or later.
Apple says CarPlay Ultra builds upon the previous version of CarPlay by providing “information for all of the driver’s screens.” That includes real-time content and gauges in the instrument cluster, “with dynamic and beautiful options for the speedometer, tachometer, fuel gauge, temperature gauge, and more.” Drivers will be able to show information from their iPhone as well, such as maps, media, advanced driver assistant systems, and tire pressure in the instrument cluster.
Netflix has more than doubled the number of people watching its ad-supported tier over the last year. At its upfront presentation for advertisers on Wednesday, the company revealed that the $7.99 per month plan now reaches more than 94 million users around the world each month – a big increase from the 40 million it reported in May 2024 and the 70 million it revealed last November.
Amy Reinhard, Netflix’s president of advertising, said the streaming service has “the most engaged audience anywhere,” with subscribers on its ad-supported tier spending an average of 41 hours per month on the service.
The new ad formats are part of Netflix’s in-house advertising platform, which is now live in the US after arriving in Canada. It plans to bring the platform to all 12 countries with ad-supported plans by June. “The foundations of our ads business are in place,” Reinhard said. “And going forward, the pace of progress will be even faster.”
Warner Bros. Discovery is changing the name of its streaming service back to HBO Max. During its Upfront event on Wednesday, the company announced that it will rebrand Max this summer, a change HBO head Casey Bloys said “better represents” its offering.
The company said in a press release that “returning the HBO brand into HBO Max will further drive the service forward and amplify the uniqueness that subscribers can expect” from the service.
“The powerful growth we have seen in our global streaming service is built around the quality of our programming,” Warner Bros. Discovery CEO David Zaslav said in a statement. “Today, we are bringing back HBO, the brand that represents the highest quality in media, to further accelerate that growth in the years ahead.”
A California judge slammed a pair of law firms for the undisclosed use of AI after he received a supplemental brief with “numerous false, inaccurate, and misleading legal citations and quotations.” In a ruling submitted last week, Judge Michael Wilner imposed $31,000 in sanctions against the law firms involved, saying “no reasonably competent attorney should out-source research and writing” to AI, as pointed out by law professors Eric Goldman and Blake Reid on Bluesky.
“I read their brief, was persuaded (or at least intrigued) by the authorities that they cited, and looked up the decisions to learn more about them – only to find that they didn’t exist,” Judge Milner writes. “That’s scary. It almost led to the scarier outcome (from my perspective) of including those bogus materials in a judicial order.”
As noted in the filing, a plaintiff’s legal representative for a civil lawsuit against State Farm used AI to generate an outline for a supplemental brief. However, this outline contained “bogus AI-generated research” when it was sent to a separate law firm, K&L Gates, which added the information to a brief. “No attorney or staff member at either firm apparently cite-checked or otherwise reviewed that research before filing the brief,” Judge Milner writes.
When Judge Milner reviewed the brief, he found that “at least two of the authorities cited do not exist at all.” After asking K&L Gates for clarification, the firm resubmitted the brief, which Judge Milner said contained “considerably more made-up citations and quotations beyond the two initial errors.” He then issued an Order to Show Cause, resulting in lawyers giving sworn statements that confirm the use of AI. The lawyer who created the outline admitted to using Google Gemini, as well as the AI legal research tools in Westlaw Precision with CoCounsel.
This isn’t the first time lawyers have been caught using AI in the courtroom. Former Trump lawyer Michael Cohen cited made-up court cases in a legal document after mistaking Google Gemini, then called Bard, as “a super-charged search engine” rather than an AI chatbot. A judge also found that lawyers suing a Colombian airline included a slew of phony cases generated by ChatGPT in their brief.
“The initial, undisclosed use of AI products to generate the first draft of the brief was flat-out wrong,” Judge Milner writes. “And sending that material to other lawyers without disclosing its sketchy AI origins realistically put those professionals in harm’s way.”
Republicans want to stop states from regulating AI. On Sunday, a Republican-led House committee submitted a budget reconciliation bill that proposes blocking states from enforcing “any law or regulation” targeting an exceptionally broad range of automated computing systems for 10 years after the law is enacted — a move that would stall efforts to regulate everything from AI chatbots to online search results.
Democrats are calling the new provision a “giant gift” to Big Tech, and organizations that promote AI oversight, like Americans for Responsible Innovation (ARI), say it could have “catastrophic consequences” for the public. It’s a gift companies like OpenAI have recently been seeking in Washington, aiming to avoid a slew of pending and active state laws. The budget reconciliation process allows lawmakers to fast-track bills related to government spending by requiring only a majority in the Senate rather than 60 votes to pass.
This bill, introduced by House Committee on Energy and Commerce Chairman Brett Guthrie (R-KY), would prevent states from imposing “legal impediments” — or restrictions to design, performance, civil liability, and documentation — on AI models and “automated decision” systems. It defines the latter category as “any computational process derived from machine learning, statistical modeling, data analytics, or artificial intelligence that issues a simplified output, including a score, classification, or recommendation, to materially influence or replace human decision making.”
That means the 10-year moratorium could extend well beyond AI. Travis Hall, the director for state engagement at the Center for Democracy & Technology, tells The Verge that the automated decision systems described in the bill “permeate digital services, from search results and mapping directions, to health diagnoses and risk analyses for sentencing decisions.”
During the 2025 legislative session, states have proposed over 500 laws that Hall says this bill could “unequivocally block.” They focus on everything from chatbot safety for minors to deepfake restrictions and disclosures for the use of AI in political ads. If the bill passes, the handful of states that have successfully passed AI laws may also see their efforts go to waste.
“The move to ban AI safeguards is a giveaway to Big Tech that will come back to bite us.”
California also came close to enacting the landmark AI safety law SB 1047, which would have imposed security restrictions and legal liability on AI companies based in the state, like OpenAI, Anthropic, Google, and Meta. OpenAI opposed the bill, saying AI regulation should take place at the federal level instead of having a “patchwork” of state laws that could make it more difficult to comply. Gov. Newsom vetoed the bill last September, and OpenAI has made it clear it wants to avoid having state laws “bogging down innovation” in the future.
With so little AI regulation at the federal level, it’s been left up to the states to decide how to deal with AI. Even before the rise of generative AI, state legislators were grappling with how to fight algorithmic discrimination — including machine learning-based systems that display race or gender bias — in areas like housing and criminal justice. Efforts to combat this, too, would likely be hampered by the Republicans’ proposal.
Democrats have slammed the provision’s inclusion in the reconciliation bill, with Rep. Jan Schakowsky (D-IL) saying the 10-year ban will “allow AI companies to ignore consumer privacy protections, let deepfakes spread, and allow companies to profile and deceive consumers using AI.” In a statement published to X, Sen. Ed Markey (D-MA) said the proposal “will lead to a Dark Age for the environment, our children, and marginalized communities.”
The nonprofit organization Americans for Responsible Innovation (ARI) compared the potential ban to the government’s failure to properly regulate social media. “Lawmakers stalled on social media safeguards for a decade and we are still dealing with the fallout,” ARI president Brad Carson said in a statement. “Now apply those same harms to technology moving as fast as AI… Ultimately, the move to ban AI safeguards is a giveaway to Big Tech that will come back to bite us.”
This provision could hit a roadblock in the Senate, as ARI notes that the Byrd rule says reconciliation bills can only focus on fiscal issues. Still, it’s troubling to see Republican lawmakers push to block oversight of a new technology that’s being integrated into almost everything.
