Aggregator

No longer content to subsume recognizable intellectual properties, the majority of the indexed internet and books (basically all of them), AI will apparently now begin devouring its own workforce.

A report in Reuters alleged that the keystrokes, mouse movements and clicks of Meta's workforce are to be captured for the purposes of training AI — something the company's communications department was happy to confirm as accurate! In a cheery missive, a company spokesperson told Engadget that "If we're building agents to help people complete everyday tasks using computers, our models need real examples of how people actually use them [...] we’re launching an internal tool that will capture these kinds of inputs on certain applications to help us train our models."

All this leads one to ask the obvious question: hey, what the fuck?

The nature of at-will employment in the United States is such that your boss basically never needs to explain why your job duties change, but it's rarely so sweeping, so brazen or so unavoidably tied to the reminder that you are being surveilled at a frighteningly granular level. Gross!

Installing keyloggers on someone else's computer in a non-work setting can often constitute a criminal offense (hello CFAA!) and it's frankly weird we allow this sort of thing to happen in the workplace at all. But in this case, there's at least some possibility this data may eventually be used to replace the exact people currently strongarmed into making those clicks and clacking those keys — or as a thin excuse to lay a lot of them off.

It's not as though the data underpinning large language models is worthless. Ill-gotten information has been the subject of exorbitant settlements and many pending court cases with considerable sums riding on their eventual judgements. If Meta thought it could obtain this sort of data from its estimated 3.5 billion combined users instead of its comparably paltry body of employees without it immediately reading as the single most invasive chapter in a laughably long history of move fast, break things, and never admit to the mess, wouldn't it just... do that? Technology has progressed so far, yet people continue to really hate feeling taken advantage of. And that sort of thing is still bad for business.

In a fragile economy floated by rampant self-dealing and the shifting moods of a few very rich weirdos, even the mere mention of AI's relentless forward march to annihilate its own creators can make a shoe company's stock pop, however briefly.

Maybe that's why Meta was delighted to confirm the broad details of the Reuters story, yet declined multiple requests to comment on if workers can opt out of this surveillance, or if they are being compensated in any way for their data. I, for one, would still love to know!

Do you work at Meta and want to talk confidentially? I'm @amarae.60 on Signal.

This article originally appeared on Engadget at https://www.engadget.com/general/hey-meta-workers-are-you-getting-paid-for-those-keystrokes-131934881.html?src=rss

BMW just announced its 7-series lineup for 2027 promising "the most extensive model update ever undertaken" by the Bavarian automaker. The series includes a pair of i7 EVs, a plug-in hybrid, two ICE models and a V8-powered M model — all running on BMW's "Neue Klasse" technology and flaunting all-new design language. 

BMW describes its updated luxury segment design language as "monolithic," touting the minimalist crystal headlights and (divisive) light-up kidney Iconic Glow grille. The body form includes a new "character line" that lends an almost "boat-tail-like" aesthetic to the three-quarter view, according to the company. BMW also hyped its new "Individual Dual-Finish" paint that pairs a matte finish in the lower area of the vehicle with a manually-applied metallic finish in the upper section for a "discrete but supremely elegant look."

The 7-series offers some shiny new interior tech as well, with an extensive upgrade for its 31.3-inch 8K BMW Theater Screen that gives back seat passengers 8K streaming, gaming and Zoom calls. If you get BMW's Digital Premium package (which includes 5G data), you'll also gain access to a TiVo powered video app, select games and the BMW Drive Recorder that gathers exterior footage in case of an accident. Naturally, it includes Apple CarPlay and Android Auto integration as standard.

The new lineup also offers special welcome and goodbye animations for when you approach and walk away from the vehicles, in Relaxed, Balanced and Excited modes. Part of that animation is the "Ceremonial Light Carpet" that projects a dynamic graphic light pattern onto the ground near the door.

With two new i7 electric sedans — the i7 50 xDrive and i7 60 xDrive — BMW has far from given up on EVs. Both promise reduced sound levels thanks to their sound-isolated electric motors, along with "instantaneous" power delivery, passenger comfort and a luxurious ambience. 

BMW's new 7 series (the gas powered version) at Grand Central. The electric version were not allowed inside the terminal,.Sam Rutherford

Both come with all-new cylindrical battery packs with 20 percent higher energy density and capacities up to 112.5 kWh. Combined with increased drive system efficiency, they'll power the BMW i7 60 xDrive up to an EPA-estimated 350 miles on a charge. You'll also be able to charge them quicker thanks to the higher 250 kW charging rate. The company claims they'll go from 10 to 80 percent in just 28 minutes with a compatible charger.

The i7 50 xDrive's dual motors offer up to 449 hp and 487 lb-ft of torque, allowing acceleration from 0 to 60 mph in 5.3 seconds with a 130 mph top speed. The i7 60 xDrive, meanwhile, packs 536 hp and 549 lb-ft of torque for a 0 to 60 mph time of just 4.6 seconds and a top speed (electronically limited) of 149 mph.

They'll also feature a new "adaptive recuperation" system that takes traffic lights into account and can automatically brake. Drivers can choose from high, medium or low brake energy recuperation, with the "high" setting offering a one-pedal feeling. BMW also introduced AI-powered "charging optimized route planning" to include charging stops if a destination is outside the vehicle's range. It can even precondition the battery to an ideal temperature to increase the charging rate as soon as the vehicle is plugged in. 

The i7 50 xDrive and i7 60 xDrive are debuting today at Auto China 2026 in Beijing and at a special New York City event at Grand Central Terminal. Production begins in July 2026 and they'll be priced starting at $106,200 and $124,700 (plus $1,550 destination and handling) respectively. 

BMW is also introducing the 750e xDrive PHEV arriving in 2027, which pairs a 308 hp six cylinder inline engine with a 194 hp electric motor for a combined peak 483 hp and 516 lb-ft of torque. No electric-only range was specified, but the top speed on all-electric power will be limited to 87 mph. That model will start production in Q4 2026, with no price yet announced. 

Finally, BMW's 740 and 740 xDrive ICE vehicles will offer up to 394 hp and 398 lb-ft of torque, offering sub-five second 0 to 60 mph sprint times and 155 mph top speeds. They'll arrive later this year at $99,800 and $102,800 respectively. All of the new 7-series models and drivetrain variants, including EVs, will be built on a single production line at BMW's Group Plant in Dingolfing.

This article originally appeared on Engadget at https://www.engadget.com/transportation/evs/bmws-new-i7-xdrive-evs-will-offer-longer-range-and-faster-charging-131059423.html?src=rss

Yoshi's solo titles have always been a product of contrasts: lovingly crafted art styles belying somewhat thin gameplay meant to appeal to a younger audience. But after getting the chance to preview Yoshi and the Mysterious Book, it feels like Nintendo has created a cozy side-scroller that retains the accessibility of earlier titles while adding extra depth and replayability for gamers of all ages. 

As you'd expect from Nintendo's lovable green dino, Yoshi's moveset consists of familiar techniques like his signature ground pound, egg toss, extendable tongue and jumping flutter kicks, the latter of which now last longer than ever before. Seriously, if you get enough height, you can float across entire levels. However, one new mechanic is that Yoshi now also possesses a tail flick that allows him to toss certain enemies onto his back. Not only is this a nice twist after years of often being treated as a beast of burden by Mario and co., it also allows Yoshi to use whoever he's carrying to interact with other elements in the stage. By flicking a Crayzee Dayzee (one of the new flower characters) into his saddle, Yoshi can make plants bloom to earn stars and uncover secrets, for example. 

Instead of selecting a stage based on its location, in Yoshi and the Mysterious Book you choose new levels by picking which character you'd like to meet next. Nintendo

Instead of simply remixing the gorgeous textile-inspired visuals we saw in Yoshi's Woolly World and Yoshi's Crafted World, Nintendo opted for a treatment that looks like a children's bedtime story. The game looks hand-drawn, while also featuring a cell-shaded effect that adds extra depth and impact to characters and backgrounds. Animations have a sort of hitch (which I initially mistook for a bug) that gives them a stop-motion effect, as if you're seeing the pages of a flipbook whiz by instead of frames being refreshed on a digital display. 

But the biggest change in Yoshi and the Mysterious Book is that it's not a traditional side-scrolling format. In many ways, this title plays like a mix between an adventure game and a platformer. Instead of just running to the right as fast as possible, the game encourages you to explore, experiment, backtrack and see what kind of new hijinks Yoshi can discover. Don't do what I did and mindlessly run around and eat up every enemy you see, because there's a good chance they are more valuable alive instead of being turned into an egg. It's almost like Nintendo applied a similar approach to subverting expectations in Super Mario Bros. Wonder, but builds on that style of gameplay even more here. 

Now Yoshi is in control of who gets to ride on his back. Nintendo

To really drive home the children's storybook conceit, Yoshi and the Mysterious Book gives you greater control over how you unlock and explore new levels. Instead of linearly moving from one stage to the next, you actually select stages by moving a magnifying glass over a new character, allowing you to learn who they are experientially. Collect enough stars from quests and you can decide where you'll go and who you'll meet next, while spending tokens found throughout each level allows you to reveal hints about how to find and complete specific objectives. Then, once you finish a stage, Mr E. (who is both the titular book and acts as the game's narrator), allows you to name each new character whatever you like. 

Instead of fighting, sometimes Yoshi just needs to help out Shy Guys by flicking them on his back and bringing them back home. Nintendo

The one thing that may rub old-school sidescroller fans the wrong way though is that Yoshi doesn't take damage — for all intents and purposes, he's invincible. This is great for kids because they don't need to worry about running out of lives or looking for coins to heal. But after playing through a couple of chapters, I found that the game can still be quite challenging — the learning process is just a bit more forgiving. One objective required me to bounce off the heads of a series of Croakaokes (the chubby amphibians that make sounds when you jump on them) to the tune of "Mary had a little lamb" while trying to reach a high-up platform. If you play a note out of order or miss a jump, it's not game over, but you will need to start over from the bottom. And even without damage, enemies like Bunchabees will still chase you around and disrupt your adventuring if you try to ignore them. 

Croakaokes love to sing when Yoshi jumps on their head. Nintendo

Yoshi's latest adventure might not be a super traditional side-scrolling platformer, but it offers a fresh take on the genre while also providing a new level of exploration and adventure for gamers of all ages. But I suppose the real test will be if it gets my own kid's seal of approval. 

Yoshi and the Mysterious Book is available for pre-order now for the Switch 2 before official sales begin on May 21 ($70 for a physical copy or $60 for a digital download).

This article originally appeared on Engadget at https://www.engadget.com/gaming/nintendo/yoshi-and-the-mysterious-book-preview-a-choose-your-own-adventure-even-adults-can-love-130000507.html?src=rss

Anker has announced its own chip that can give its small, wearable products AI capabilities that run locally on device. The company is planning to debut the chip called “Thus” on a new model of headphones, slated to be unveiled at its Anker Day event on May 21.

Anker calls Thus the “first Compute-in-Memory (CIM) AI audio chip with neural networks.” The company explains that Thus is “inspired by the workings of the human brain” in that the storage and processing of information takes place in one location instead of keeping them separate, similar to how it works on modern chips for computers.

Thus integrates computing power directly into NOR flash memory cells, which provide faster read speeds than NAND memory. A NOR-based CIM system requires only a tiny space inside devices, which makes it an ideal option for small products like headphones. Anker says headphones are a particularly challenging environment to demonstrate what a new chip can do, because “hardly any other device places higher demands on an AI chip.” They have a tiny space allotted for components and operate with just a few milliwatts of power, even though they have to consistently provide noise cancellation. If the model delivers, it could be a huge advertisement for Thus, which Anker plans to put in other mobile accessories and IoT devices, as well.

While the company has yet to reveal all its upcoming headphones’ AI-powered capabilities, it did announce one particular feature. Clear Calls, as it’s called, will cancel noise “with a large neural network running entirely on the device, supported by eight MEMS microphones and two bone conduction sensors.” Anker says it will enable significantly clearer conversations even in environments that are challenging for conventional noise cancellation.

This article originally appeared on Engadget at https://www.engadget.com/ai/ankers-thus-chip-brings-ai-to-its-headphones-and-other-products-122142552.html?src=rss

Nikita Bier, X’s head of product, has announced the launch of custom timelines, which lets you curate what you see on your feed based on your topics of interest. He called the update “one of the biggest changes to X” and a ”huge undertaking” that took the team “many months” to develop. The feature lets you pin specific topics to your home tab, so you can switch from one to the other to see the latest discussions about your interests and hobbies.

Bier said that X’s custom timelines is “powered by Grok's understanding of every post with the algorithm's personalization.” You have 75 topics to choose from, including food, art, photography, business, finance, movies and TV. As you’d expect, the personalization aspect of the feature works better if it’s a topic you already engage with regularly. X’s new feature is similar to Bluesky’s and Threads’ custom feeds, which also allow you to pin topic-based timelines to the home screens of the apps, and which their users have been enjoying since 2023 and 2024, respectively.

At the moment, X’s custom timelines is still in its early access phase and is only available to Premium subscribers on iOS. It will be rolling out to Premium users on Android “very soon,” as well. Bier has also announced that X has released a tool to snooze topics on the For You tab. With the tool, you’ll be able to hide certain topics, such as politics or sports, for 24 hours from your feed. It’s now available for Premium users on iOS and the web.

Ladies and gentlemen, today we're launching one of our biggest changes to 𝕏

Introducing Custom Timelines

This feature allows you to pin a specific topic to your home tab. With support for over 75 topics, you can dive deep into your favorite niche on X.

It's powered by Grok's… pic.twitter.com/9jkIEXvubj

— Nikita Bier (@nikitabier) April 21, 2026

This article originally appeared on Engadget at https://www.engadget.com/social-media/x-finally-adds-custom-timelines-103130966.html?src=rss

Anthropic is investigating potential "unauthorized access" to its Claude Mythos model that has been touted for its ability to find cybersecurity flaws, the company told Bloomberg. A group gained access to the model through a third-party contractor portal and by using internet sleuthing tools, according to the report. However, the group is only interested in trying the models and not using them maliciously, according to a person familiar with the matter. 

"We're investigating a report claiming unauthorized access to Claude Mythos Previous through one of our third-party vendor environments," Anthropic said in a statement. 

The Claude Mythos Preview arrived earlier this month as part of "Project Glasswing" with significant fanfare. Anthropic limited the preview release to a small number of trusted test companies including Amazon, Microsoft, Apple and Cisco. Another was Mozilla, which said the model helped it find and patch 271 Firefox vulnerabilities. A growing number of banks and government agencies have been seeking access as well in order to safeguard their own systems. 

However, several unauthorized users (who reportedly have a private chat on Discord), supposedly gained access to Mythos through a developer portal and by making an educated guess as to where the model might be located. That same group may also have access to other unreleased Anthropic models, according to the report. 

The new Mythos model has gained notoriety of late for its supposed ability to sniff out security flaws in operating systems and internet browsers. This has prompted some skepticism among security researchers but also fear that AI-generated cyber attacks could become a "real threat," CTO of cloud security firm Edera Alex Zenla recently told Wired. Anthropic was recently designated as a "supply chain risk" by the US Department of Defense, but has been in talks with the Trump administration of late to have that label removed. 

This article originally appeared on Engadget at https://www.engadget.com/ai/anthropic-is-investigating-unauthorized-access-of-its-mythos-cybersecurity-tool-091017168.html?src=rss

SpaceX and AI company Cursor have struck a new partnership that could see the owner of X buy the AI company for $60 billion later this year. "SpaceXAI and  @cursor_ai  are now working closely together to create the world’s best coding and knowledge work AI," SpaceX wrote in a post on X. 

SpaceXAI and @cursor_ai are now working closely together to create the world’s best coding and knowledge work AI.

The combination of Cursor’s leading product and distribution to expert software engineers with SpaceX’s million H100 equivalent Colossus training supercomputer will…

— SpaceX (@SpaceX) April 21, 2026

According to SpaceX, the deal allows for it to either invest $10 billion into the company known for its AI coding tool, or acquire it entirely "later this year" for $60 billion. If an acquisition were to happen, it's not clear at what point Cursor could officially join the fold of Elon Musk's rapidly expanding and increasingly enmeshed web of companies. SpaceX bought xAI, the billionaire's AI company that also controls X, earlier this year. SpaceX is currently getting ready to go public this summer in what will likely be the biggest initial public offering (IPO) in history. 

Cursor, which has reportedly been in talks to raise its own $2 billion round of funding, is known for its AI coding tool of the same name that's become the vibe coding platform of choice for many developers. It allows people to use either its own models or those from other leading AI companies, including OpenAI, Google, Anthropic and xAI.

In a statement, Cursor said its partnership with SpaceX will "accelerate our model training efforts" while addressing infrastructure-related issues that have slowed it down in the past. "We've wanted to push our training efforts much further, but we've been bottlenecked by compute," the company said. "With this partnership, our team will leverage xAI's Colossus infrastructure to dramatically scale up the intelligence of our models for coding and beyond."

This article originally appeared on Engadget at https://www.engadget.com/ai/spacex-and-cursor-strike-partnership-that-might-end-in-a-60-billion-acquisition-232131487.html?src=rss

Anthropic's buzzy announcement about using AI to improve cybersecurity earlier this month was met with plenty of skepticism. However, Mozilla shared some details that support use of the company's special Claude Mythos Preview model as a way to protect critical services. Using Mythos helped Mozilla's team find and patch 271 vulnerabilities in the latest release of the Firefox browser. "So far we’ve found no category or complexity of vulnerability that humans can find that this model can’t," the foundation said.

The blog post from Mozilla feels like a positive sign for Anthropic's Project Glasswing. Obviously the AI company would want to put itself in the best possible light while presenting its own initiative, but there's something encouraging about hearing the benefits from a third party. Mozilla also noted that in its time with Claude Mythos, the AI wasn't able to turn up any bugs that a human wouldn't have been able to find, given enough time and resources, which indicates that AI isn't presently able to do more to crack cybersecurity protections than a person can.

An organizaion successfully using AI for good is certainly a refreshing change of pace in tech news. And for those Firefox users who aren't personally interested in applying any generative AI in their browsing, Mozilla has given the option to turn it all off for the past several months.

This article originally appeared on Engadget at https://www.engadget.com/ai/mozilla-says-it-patched-271-firefox-vulnerabilities-thanks-to-anthropics-claude-mythos-224330023.html?src=rss

Cash App, the banking and payments app run by Block, has added support for parent-managed kids accounts. The new accounts include key benefits from the service's normal account, with an eye towards teaching financial literacy to younger users ages 6 to 12. Cash App first allowed teenage users on its platform in 2021.

As part of the "expanded Cash App Families experience," eligible legal guardians and parents can create managed accounts that offer "a dedicated place on the platform to send allowances, set aside savings, and track spending for their child, kickstarting their path to financial independence," Cash App says. Adults managing these accounts will be able to set up recurring transfers, see how their child is spending and do things like lock their child's account to prevent transactions. Kids will get a custom debit card and the ability to receive payments from up to five trusted accounts, though notably they won't be able to access Cash App itself.

Today, we're launching Cash App accounts for kids age 6-12. Parents manage the accounts. Kids get to learn about safety, start saving for goals, and design and use their own debit card.

Next generation banking never looked so good.

Proud of the team for this one. pic.twitter.com/jIAcbvsfB9

— Kristen Anderson (@FintechKristen) April 21, 2026

Cash App says managed accounts are designed for kids 6 through 12. Once those kids turn 13, Cash App says parents will be able to choose to convert their account to a "sponsored account" to unlock more features, like the ability to send and receive payments, invest in stocks or trade crypto. Those sponsored accounts are technically still monitored and controlled by a parent or legal guardian, but they do give 13-year-olds more control over how they use their money.

A parent-managed account for kids is not a new idea in the fintech space, though Cash App is trying to reach a younger audience than some of its competitors. Venmo rolled out access to its payment platform to teens between the ages of 13 to 17 in 2023. Separately, both Apple and Google also offer their own kids accounts in Google Wallet and Apple Cash Family.

This article originally appeared on Engadget at https://www.engadget.com/apps/cash-app-now-supports-accounts-for-kids-6-12-210651025.html?src=rss

YouTube notifications can get messy fast, particularly if you’re subscribed to a lot of different channels. To address that, today the company will begin muting push notifications from creators that you haven’t engaged with in the last month.

The change to YouTube notifications began as a small trial the company tested out earlier this year. The idea behind it is that if a viewer continually receives notifications about content they don't engage with, this may eventually cause the user to disable YouTube notifications altogether. Now obviously, this is bad for YouTube. Turning off notifications means people will use the platform less, thereby resulting in lower revenue. However, it's also bad for content creators, especially the ones you do like, who will have one fewer avenue to keep you updated about new and upcoming videos. 

So starting today, for channels that you have subscribed to and have notifications set to "all," YouTube will no longer send out push notifications to mobile devices from creators that you haven't interacted with for one month. That said, these notifications will continue to be available inside the YouTube app in your inbox (the little bell icon in the top right). 

Notably, for those who are clicking on notifications and watching related videos, nothing will change. Additionally, based on info from the test earlier this year, YouTube said "channels that upload infrequently will not have their notifications affected." This is a good thing, especially for creators who post long-form content that takes extra time to make, as people probably don't want notifications to go away in case they happen to miss a once-a-month upload. 

The one thing that's unclear is if you start watching a channel again that you have not interacted with in a while, is if YouTube will automatically restart related push notifications. However, as a way to prevent too many alerts from clogging up your phone, YouTube's new protocol seems like a good way to cut down on the clutter.

This article originally appeared on Engadget at https://www.engadget.com/entertainment/youtube/youtube-is-muting-push-notifications-from-channels-you-dont-watch-205119228.html?src=rss

When online platforms violate their own privacy policies to sell your photos, have no fear: They just might have to pay an undisclosed settlement fee 12 years later. (Who says justice is dead?) According to Reuters, AI company Clarifai says it has deleted 3 million profile photos taken from dating site OkCupid in 2014. It follows a settlement reached last month between the FTC and Match Group, OkCupid's owner.

The Delaware-based Clarifai reportedly certified the data deletion to the FTC on April 7. The company also confirmed to US Representative Lori Trahan (D-MA) that it deleted any models that trained on the data. Clarifai told the representative's office that it hadn't shared the data with third parties.

The FTC opened the investigation in 2019, after The New York Times reported that Clarifai had built a training database using OkCupid dating profile photos. The behavior was a direct violation of OkCupid’s privacy policy. Court documents reviewed by Reuters reveal that Clarifai asked OkCupid executives for the data in 2014. Apparently, they obliged.

<p>Clarifai uses this creepy facial profiling example to sell its services.</p>Clarifai

"We're ⁠collecting data now and just realized that OkCupid must have a HUGE amount of awesome data for this," Clarifai founder Matthew Zeiler wrote in an email to OkCupid co-founder Maxwell Krohn. The AI startup used the dating site's images to build a facial recognition service that can identify a person's age, gender and race. (Another brilliant and totally ethical idea from Clarifai, tapping into unsecured city surveillance cameras without authorization, was reportedly shuttered.)

Zeiller suggested to The New York Times in 2019 that people needed to, well, get over it. "There has to be some level of trust with tech companies like Clarifai to put powerful technology to good use, and get comfortable with that," the AI founder declared. Some of OkCupid's founders were reportedly investors in Clarifai.

As part of the settlement, the FTC "permanently prohibited" OkCupid from misrepresenting its data collection and privacy controls. TechCrunch notes how strange it is to use that as a penalty, given that FTC rules already bar that behavior.

This article originally appeared on Engadget at https://www.engadget.com/ai/ai-company-deletes-the-3-million-okcupid-photos-it-used-for-facial-recognition-training-195223996.html?src=rss

Meta is facing a new lawsuit over its advertising practices. The nonprofit group Consumer Federation of America (CFA) has filed a proposed class-action suit against Meta for "failing to protect users" from scam ads on Facebook and Instagram. 

The lawsuit, which was first reported by Wired, alleges that Meta has run afoul of consumer protection laws in Washington D.C. for misleading Facebook and Instagram users about scams on its apps and that the company has "chased profits rather than protecting its users." The filing includes numerous examples of alleged scam ads that CFA says it found in Meta's ad library. These include ads promoting a "free government iPhone," as well as those claiming to offer $1,400 checks to people born in certain years. Many of the ads use AI videos, according to CFA.  

Some of examples of alleged scam ads CFA includes in its lawsuit.CFA

Meta's advertising practices have been in the spotlight since last year when Reuters reported on internal documents that indicated the company was making billions of dollars from ads promoting scams and banned goods. The report also highlighted how Meta's own processes have at times made it harder for its own employees to fight malicious advertisers.

"Meta claims it is doing all it can to crack down on scam advertising on its platforms," CFA's lawsuit states. "But in reality, Meta has knowingly taken steps and adopted policies that pad its bottom line at the expense of its users’ safety and well-being. In fact, rather than prohibiting advertisers who the company itself has determined pose a higher risk to its users (as other tech companies like Google have), Meta just charges these advertisers more. The perverse result is that the riskier the advertiser, the more money Meta makes."

CFA's allegations "misrepresent the reality of our work and we will fight them," a Meta spokesperson said in a statement. "We aggressively combat scams across our platforms to protect people and businesses — last year alone, we removed over 159 million scam ads, 92% of which we took down before anyone reported them, and took down 10.9 million accounts on Facebook and Instagram associated with criminal scam centers. We fight scams because they are bad for business — people don't want them, advertisers don't want them, and we don't want them either.”

This article originally appeared on Engadget at https://www.engadget.com/social-media/meta-has-misled-users-about-scam-ads-on-facebook-and-instagram-lawsuit-says-193220235.html?src=rss

New York is the latest state to take a stand against prediction markets. Attorney General Letitia James has sued Coinbase Financial Markets and Gemini Titan on charges that both are illegally running unlicensed gambling operations. The suit also claims that these prediction markets violate state laws that prevent betting on games involving New York college sports teams. 

"Gambling by another name is still gambling, and it is not exempt from regulation under our state laws and Constitution," James said. "Gemini and Coinbase’s so-called prediction markets are just illegal gambling operations, exposing young people to addictive platforms that lack the necessary guardrails."

Multiple states have taken similar actions over the proliferation of prediction markets, but they may face a new roadblock at the federal level. Earlier this month, the US Commodity Futures Trading Commission sued three of the states that have charged prediction markets with running unlicensed gambling. The CFTC claimed that it should be the sole regulator for prediction markets and called the efforts by Arizona, Connecticut and Illinois an overreach of authority.

This article originally appeared on Engadget at https://www.engadget.com/big-tech/new-york-attorney-general-sues-two-prediction-markets-on-illegal-gambling-allegations-192012225.html?src=rss

Florida Attorney General James Ulthmeier has announced that the state's Office of Statewide Prosecution has opened a criminal investigation into OpenAI and ChatGPT. The investigation was opened because the suspect in a mass shooting at Florida State University in 2025 reportedly used ChatGPT in the lead up to the shooting.

Per Uthmeier, "Florida law states that anyone who aids, abets, or counsels someone in the commission of a crime, and that crime is committed or attempted, may be considered a principal to the crime." That means that the responses provided by ChatGPT to the shooter could be interpreted as the AI assistant aiding and abetting his actions. Or at least that's what Florida seems interested in arguing.

OpenAI provided the following statement when asked to comment on the Florida investigation:

Last year's mass shooting at Florida State University was a tragedy, but ChatGPT is not responsible for this terrible crime. After learning of the incident, we identified a ChatGPT account believed to be associated with the suspect and proactively shared this information with law enforcement. We continue to cooperate with authorities. In this case, ChatGPT provided factual responses to questions with information that could be found broadly across public sources on the internet, and it did not encourage or promote illegal or harmful activity. ChatGPT is a general-purpose tool used by hundreds of millions of people every day for legitimate purposes. We work continuously to strengthen our safeguards to detect harmful intent, limit misuse, and respond appropriately when safety risks arise.

As part of the investigation, Florida has subpoenaed OpenAI for information on "all policies and internal training materials" related to how the company handles things like users threatening to harm others, threatening to harm themselves and how OpenAI responds to law enforcement. The state is also asking OpenAI to share its organizational chart and any publicly released statements on the shooting.

"Florida is leading the way in cracking down on AI's use in criminal behavior, and if ChatGPT were a person, it would be facing charges for murder," Attorney General James Uthmeier said. "This criminal investigation will determine whether OpenAI bears criminal responsibility for ChatGPT's actions in the shooting at Florida State University last year."

Florida’s investigation isn’t the first time OpenAI has been connected to a mass shooting. Canadian regulators called for OpenAI to change how it approaches threats of harm following a Wall Street Journal report that claimed the company flagged the account of a Canadian shooting suspect in 2025 but failed to bring their threats to law enforcement. The company agreed to new policies around how it works with Canadian law enforcement in March. Separately, OpenAI is still in the midst of a wrongful death lawsuit from 2025 for the role it may have played in the suicide of a teenage user.

This article originally appeared on Engadget at https://www.engadget.com/ai/florida-ag-opens-criminal-investigation-into-openai-and-chatgpt-190200227.html?src=rss

A little more than a year after OpenAI gave ChatGPT users the option to create images and designs directly from its chatbot, it's now releasing ChatGPT Images 2.0. OpenAI describes the new system as a “step change” for image generation models, particularly when it comes to the tool’s ability to follow instructions in detail, render dense text and place and relate objects in a scene. For the first time, OpenAI has also built an image model with reasoning capabilities, giving the system the ability to do things like search the web and verify its outputs. According to the company, those capabilities should translate to a tool that's more reliable when accuracy, consistency and visual cohesion are essential. 

An example of ChatGPT's new non-Latin rendering abilities. OpenAI

OpenAI says it has also put in a lot of work to make Images 2.0 better at understanding and rendering non-Latin text, with "significant gains" when it comes to the model's ability to handle Japanese, Korean, Chinese, Hindi and Bengali. At the same time, the company claims the new model is better at faithfully recreating the specific characteristics of different visual languages. On this point, OpenAI says that makes Images 2.0 more useful for tasks like game prototyping and storyboarding. Outside of those features, the new model is more flexible when it comes to aspect ratios, allowing it to generate images that are as wide as 3:1 and as tall as 1:3. It can also produce designs at resolutions of up to 2K, and even generate up to eight outputs in one go. 

A tortoiseshell cat in the style of Pokemon's third generation of games. ChatGPT

I got a chance to preview Images 2.0 ahead of its public release. For my first prompt, I asked ChatGPT to generate an image of a tortoiseshell cat in the pixel art style of Pokémon's third generation. I thought this would be a good test because AI models typically struggle with pixel art, and the Game Boy Advance Pokémon games are iconic for their art style, so much so that if ChatGPT merely approximated that style, it wouldn't do. The result is the image you see above, and I think ChatGPT did a commendable job there. I then tasked the new model with converting that image into a transparent PNG. For one last test, I asked ChatGPT to create a four-page manga about my cat enjoying a sunny day by an idyllic city stream. 

Notice how the cat isn't render exactly like the one above it. ChatGPT

Of those three tests, ChatGPT spent the most time on the second one and the output there was slightly different from the first image it generated, which I felt deviated from my prompt. Still, it managed to generate a proper transparent image, which is something other image models can struggle to do properly. Once more people have a chance to put the model through its paces, we’ll have a better idea of how it compares to Google’s Nano Banana 2, and where OpenAI can make additional improvements.

A manga generated by ChatGPT about a cat enjoying a sunny day. ChatGPT

Images 2.0 is available starting today for all ChatGPT users, including those on the company's Free and Go tiers. Plus and Pro subscribers get access to more advanced outputs. OpenAI is also making the model available through its API service and Codex coding app, which just last week it updated to offer built-in image generation. Notably, Images 2.0 arrives just days after Anthropic waded into the visual design market with its own design assistant.  

This article originally appeared on Engadget at https://www.engadget.com/ai/chatgpt-images-20-is-better-at-rendering-non-latin-text-190000153.html?src=rss

Netflix's hit animated adaptation of the video game series Devil May Cry is returning for a second season on May 12. The streamer has released a full trailer alongside this announcement, after dropping a short teaser several months back.

The new footage promises plenty of action and franchise fans will note that protagonist Dante now looks extremely similar to his counterpart from the games. This tracks with what series creator Adi Shankar said last year, when he promised that season two would show the character "embrace more of the iconic badassery fans of the game expect."

The show, and the games, follow a half-demon demon hunter as he fights to prevent the gates of Hell from opening. The first season was a gigantic hit for the platform, racking up more than 5 million views in its first four days of release.

Showrunner Shankar was heavily involved with the animated Castlevania adaptation, also for Netflix. Studio Mir is once again handling the animation, which is always great to see. This is the studio behind The Legend of Korra, X-Men '97, Kipo and the Age of Wonderbeasts and My Adventures with Superman, among others. The first season of Devil May Cry looked gorgeous, so season two should follow suit.

This article originally appeared on Engadget at https://www.engadget.com/entertainment/tv-movies/netflixs-devil-may-cry-animated-adaptation-returns-for-a-second-season-on-may-12-184003226.html?src=rss

Framework’s Laptop 13 Pro may be the star of today’s launch, but it is by no means the only product the company is unveiling. We’re also seeing some quality-of-life upgrades for the Laptop 16, a new wireless keyboard and a carrying case. Plus, it’s addressing something a vast number of its own users have demanded for a while: A 10GB Ethernet expansion card.

Laptop 16 Framework

The top deck of the Laptop 16 is sufficiently customizable that you can cram in any number of modules alongside the keyboard and trackpad. If you want an area blank, then you can just put in the requisite spacer, but that creates an unintended problem for some users. CEO Nirav Patel admitted the areas where the modules sit side-by-side can trap some people’s arm hairs, leading to some unplanned depilations when they move away from the deck. To solve this, the company is releasing a new all-in-one trackpad and all-in-one keyboard cover. The trackpad has also been updated to the same haptic model found in the 13 Pro. Plus, it’s now available with a lower-end Ryzen 5 340 mainboard as a more affordable entry level model.

OCuLink Dev kit

At the back of the Laptop 16 is an expansion module slot that either holds a blank unit or the discrete graphics hardware. Last year, the company launched a module with an RTX 5070 inside and, now, it’s started thinking about using it for eGPUs as well. The OCuLink Dev Kit (pictured, above) is a prototype board using OCuLink 8i which can connect directly to your mainboard over its native PCIe lanes.

If you already own a blank module, you won’t need to buy a replacement either. Framework has ensured you can simply drop the OCuLink board into the slots of your existing one. As well as the board, the kit includes an OCuLink Dock that’ll let you slot in any off-the-shelf PCIe card — including GPUs, 100Gb Ethernet and video capture — and run it as if you had your very own desktop.

Sadly, it’s not yet ready for prime time and so far all the company has said is we’ll see more about the hardware “later this year.”

Wireless Touchpad Keyboard Framework

“There’s one living room keyboard that everyone owns, that Logitech keyboard they haven’t updated in 15 years,” said Patel. “I’ve got two of them, polled the team, and a bunch of them all have the same keyboard,” he added. But Patel’s frustration drove him and his team to build their own, better model, which has been an idea since development work started on the Framework Desktop. It was only when the team learned that the Desktop was being used as a living room PC, however, that they actually brought it to market. It is, in fact, using the same excellent keyboard hardware as found on the Framework Laptop 12, repackaged for its all-in-one form.

Laptop Sleeve Framework

Framework’s built a laptop sleeve containing extra pockets for your spare expansion cards and screwdriver. Patel said that the real focus was on ensuring the materials used had as small a carbon footprint as possible.

10Gb Ethernet Expansion Card Framework / WisdPi

Rounding out the announcements is a 10Gb Ethernet expansion card for users who want a faster wired internet experience than the current 2.5Gb model. But this isn’t a Framework product per-se, but was designed and developed by hardware makers Wisdpi. Patel said the community is so passionate that they’re constantly looking at new chip and hardware announcements. They then head to the company’s forums to ask “Hey, like, can we get a Framework version of this?” he explained. The advent of a slender 10Gb Ethernet chip was enough to inspire one third-party developer to build an Expansion Card module to harness it. “Basically, we got in contact with them,” said Patel and offered some guidance to help build the module. And once it was finished, Framework is now acting as a reseller to enable the whole community to easily buy it. Patel admitted that a 10Gb Ethernet Card would inevitably cater to a fairly niche customer, but “love to see a third party come in and fill that gap.”

This article originally appeared on Engadget at https://www.engadget.com/computing/laptops/framework-is-building-an-egpu-kit-for-its-laptop-16-181516238.html?src=rss

Framework’s lineup of modular, repairable laptops has seen the company grow from a niche to the mainstream. Now, the company is launching a pro version of its 13-inch laptop, but it’s still held close to all of its principles. Because while this all-new version has plenty of bells and whistles, almost all of its components are still instantly compatible with the rest of the range. You can take a part from this new 13 Pro, and install it into the first-generation 13 launched back in 2021 without much fuss.

Framework Laptop 13 Pro is touted as a “ground up redesign” of the existing 13, taking into account feedback from its dedicated and passionate users. That includes a far bigger battery, new chassis, new memory, haptic trackpad and a custom touch display. It also comes in black and, even in the press images, it’s immediately clear it’s a better color for the company’s austere industrial design. CEO Nirav Patel smiled knowingly when I said it’s immediately evocative of a ThinkPad, and I mean that in the most complimentary way.

Framework

The biggest change has been to boost the battery to 74Wh to address gripes about longevity. It’s the second time Framework has boosted the cell size, which started at 55Wh and presently runs to 61Wh. To make the battery fit, the bottom of the chassis has been redesigned, filling out the chamfers present on the existing 13. Framework says the Pro’s lifespan will hit 20 hours of uptime while streaming Netflix in 4K, and says it’ll post the videos to YouTube to prove it.

Given the redesigned lower chassis, the new battery is the one part you can’t simply drop into an older machine. “You’ll need the new bottom cover to fit,” explained Patel, “but because we’ve also increased the thickness of the battery, you have to switch over to the new input cover that has the haptic trackpad.” Patel added while you may need to pair up some parts from upgrades, there’s no component that you can’t retrofit. The new chassis means the speakers are now side-firing, and are now Dolby Atmos-certified.

The bigger battery is working in tandem with Intel’s new Core Ultra Series 3 chips which promise to be incredibly efficient. Both Intel and Framework are sure the Panther Lake silicon is going to sip at that beefy battery, but with enough grunt to play AAA games. Users will get the pick of a Core Ultra 5, X7 or X9, with the promise all of them will be able to eat a game like Cyberpunk 2077 for breakfast, lunch and dinner. At the same time, the Pro will also launch with an AMD Ryzen AI 300 series mainboard option, which are the same mainboards found on the 2025 Laptop 13. 

Framework

The last major shift has been in the RAM, going from regular SO-DIMMs to LPCAMM2. The newer hardware design enables the use of LPDDR5X RAM with its better power efficiency and faster memory bandwidth. And it’s not soldered to the board, so you can add in more at some unspecified future date. Which is important if, I dunno, for some reason the global RAM market suddenly crunches and you need to mortgage a kidney for a single stick.

Naturally, I wanted to know how this would impact the laptop’s thermals, Framework’s weak spot. Patel said the mainboard’s cooling has been tweaked (once again), this time with some extra help from Intel. He added the Pro he was using to run our call hadn’t spun up its fans in half an hour which, for a Framework, is a big deal. As always, I’ll reserve judgment until I’ve seen the thing in person.

The Pro is also a way for Patel to fix some of the compromises the company had to make at its genesis. “As a startup six years ago, we just didn’t have access,” said Patel, “when we go to a display vendor [now] and say ‘we want a custom panel’ they say ‘okay, let’s talk specs’ rather than ‘who are you?”’ The 13 Pro ships with a 13.5-inch, 3:2, 2,880 x 1,920 touch display with a variable refresh rate between 30 and 120Hz. Its backlight pushes all the way to 700nits, it has a 1800:1 contrast ratio and per-unit color calibration.

I was surprised the 13 Pro has touch given it’s still a 13-inch productivity notebook without a wraparound hinge. “All the way back in 2021, when we first launched, we were getting asked for touch support,” said Patel. But the company was naturally resistant, assuming touch was an unnecessary addition for a machine of this class. It was only when building the Laptop 12 did the team find it to be quite useful. “It was a slam dunk. We know people want it, we used it on the 12, we like it and it doesn’t actually cost that much. Let’s just add it in,” said Patel.

Framework

Framework knows touchpads are a weak spot for Windows laptops when compared to those made by Apple. Consequently, the Pro 13 has a haptic touchpad with four piezo elements that, it’s hoped, will stand shoulder-to-shoulder with a MacBook Pro. The new input cover keeps the came keyboard and fingerprint sensor, wisely not touching a feature that isn’t broken and doesn’t need fixing. What is cool, however, is the more punky gray and black aesthetic on the keyboard that splits the difference between a ThinkPad and a cyberdeck. “A big part of this goes back to the idea to build the ultimate developer laptop,” said Patel.

Existing Framework owners may be concerned about how this will affect their machine’s longevity. The company might not like me saying, but to me, the Pro name is only meaningful if you’re buying a new machine off the shelf with the new features in one package. There’s nothing stopping you from upgrading your existing machine to match the pro in one go, or as your needs evolve. That’s fundamentally the company’s greatest strength, since it has committed to bringing every single one of its existing users along. But doesn’t make for the splashiest headlines when it comes to debut a new product.

Framework Laptop 13 Pro is available to pre-order today, with the first shipments due to start in June. The base-model pre-built Windows system will start at $1,699, while the DIY model will set you back $1,199.

This article originally appeared on Engadget at https://www.engadget.com/computing/laptops/framework-launches-the-laptop-13-pro-with-intels-new-panther-lake-chips-181503934.html?src=rss

Monterey Park's city council has moved to ban construction of any data centers within its borders. The California city's leaders placed a permanent ban on these buildings, labelling them a public nuisance. A proposed plan to construct a 250,000 square foot data center was stopped after residents and advocates pushed back against the project. 

Tech journalist Brian Merchant reported on the public comment phase of the city council meeting where residents spoke decisively about data centers. "I can tell you that this issue has brought left, right and center together. It’s a quality of life issue," one commenter said. "Don’t let the rich steal our future."

Monterey Park may be the first US city to lay down the law blocking data center projects, but others are primed to follow suit. New York's state leadership is working on legislation that would prevent data center construction for three years. Maine has a similar bill that has already made it to the governor's desk. At the federal level, Rep. Alexandria Ocasio-Cortez (D-NY) and Senator Bernie Sanders (I-VT) have proposed a ban on building new data centers until there are more guardrails in place for AI development and environmental security. 

Existing facilities have also faced some pushback. For instance, the NAACP is suing xAI for alleged violations of the Clean Air Act at its data center in South Memphis.

This article originally appeared on Engadget at https://www.engadget.com/ai/monterey-park-california-has-banned-any-data-centers-within-its-city-limits-180426656.html?src=rss

The search for signs of life on Mars continues to yield promising data. A first-of-its-kind wet chemistry experiment, published Tuesday in Nature, confirmed the presence of essential ingredients of life preserved in ancient Martian sandstones.

The molecules were found inside 3.5-billion-year-old sandstone. NASA's Curiosity rover collected the clay-filled rocks from an area called Glen Torridon, inside Mars' enormous Gale Crater. The rover's Sample Analysis at Mars (SAM) mobile instrument suite analyzed the data.

The experiment was unique as the first off-Earth study to use the chemical tetramethylammonium hydroxide (TMAH). The reagent allows Curiosity to break down larger organic molecules on the Martian surface, reducing them to something the rover's instruments can read.

It revealed the presence of over 20 different organic molecules. Among the data was confirmation of naphthalene and benzothiophene, some of the largest and most complex organic compounds discovered on the Red Planet. The experiment also yielded the first detection of a possible N-heterocycles, which DNA and RNA are built upon.

<p></p>NASA

“That detection is pretty profound because these structures can be chemical precursors to more complex nitrogen-bearing molecules,” the paper’s lead author, Amy Williams, wrote in NASA’s announcement. “Nitrogen heterocycles have never been found before on the Martian surface or confirmed in Martian meteorites.”

As with previous discoveries of organic material on Mars, this one is not yet the smoking gun we've been waiting for. But it adds to a growing body of evidence that, at a minimum, the foundations of life as we know it were present on an ancient version of the planet. The study also confirms that organic material can survive on Mars for billions of years, which will encourage future experiments.

The paper's authors say the data will help NASA to optimize its second (and final) TMAH experiment on Curiosity. It also opens the door to future TMAH tests on the Rosalind Franklin Mars rover and the Dragonfly mission to Saturn's moon, Titan. Both missions are scheduled for 2028 at the earliest.

This article originally appeared on Engadget at https://www.engadget.com/science/space/nasas-curiosity-rover-found-promising-organic-chemicals-on-mars-174514375.html?src=rss