Apple is scrambling to make software updates to Apple Watch algorithms that measure blood oxygen levels to avoid an impending ban on the smartwatch in the US over a patent dispute, Bloomberg reported. Changing how the Watch measures oxygen saturation, Apple believes, could help keep the Watch on shelves during the busy holiday shopping season.

The blood oxygen sensor, which was first introduced with the Apple Watch 6 in 2020, is at the heart of a patent dispute between Apple and Masimo, another California-based company that sued Apple in 2021. Masimo claimed that Apple’s sensor violated two patents related to light-based blood-oxygen monitoring that it owned. In October, the International Trade Commission (ITC) upheld a ruling, stating that Apple did, in fact, violate Masimo’s patents.

The case then went to the White House for a 60-day Presidential Review period, which ends next week. If President Biden doesn’t veto the ITC’s decision, Apple will be banned from selling the Apple Watch 9 and Apple Watch Ultra 2, which include blood oxygen sensing. Apple is complying preemptively with the ban and will stop selling both Apple Watch models on its website on December 21 and in retail stores on December 24 in case the veto doesn’t happen.

One way Apple could keep selling the watch is to settle with Masimo, but the company’s last-minute race to make software changes to the Watch suggests that it doesn’t plan to do so. The company told Bloomberg that it plans to submit its software workaround to the ITC for approval. Masimo’s CEO Joe Kiani, told Bloomberg that he would be open to settling with Apple, but the company hasn’t called him yet. “It takes two to tango,” he said.

He also doesn’t think that a software update to the Apple Watch would resolve the situation. “I don’t think that could work — it shouldn’t — because our patents are not about the software,” Kiani said. “They are about the hardware with the software.”

Overhauling Apple Watch hardware would, of course, be a lot more challenging for Apple than tweaking its software. Even if the ITC approves any potential hardware changes, manufacturing and shipping modified versions of the two Watch models could take upwards of three months, a person familiar with how Apple operates told the publication.

Apple and Masimo have a complicated history. Apple reportedly discussed acquiring the company in 2013 and hiring Kiani to work on the medical features on its smartwatch. That deal never went through. Instead, Kiani claimed that Apple hired more than 20 Masimo engineers, doubled some of their salaries, and made them develop the same kind of medical technology they were working on at Masimo at Apple. “This is not an accidental infringement — this is a deliberate taking of our intellectual property,” Kiani told Bloomberg.

Engadget has reached out to Apple for comment on Kiani’s statements. The company has previously called the ITC’s ruling “erroneous” and plans to appeal the decision to the Federal Circuit.

It’s not clear whether Apple will get immediate relief. The silver lining for the company is that the ban only applies to Apple Watch sales through the company’s own channels — its website and its retail stores — in the US. You should still be able to buy the Apple Watch through other retailers like Best Buy, Walmart and Target.

OpenAI will pay German publisher Axel Springer to use its news articles to train its AI models and show real-time information from Axel Springer's brands, which include Business Insider and Politico in the US and Bild and Welt in Europe, in ChatGPT’s responses. None of the companies disclosed how much the deal was worth, but Bloomberg reported that OpenAI will pay the publisher tens of millions of euros over the next three years.

“This partnership with Axel Springer will help provide people with new ways to access quality, real-time news content through our AI tools,” said OpenAI’s chief operating officer Brad Lightcap in a statement. “We are deeply committed to working with publishers and creators around the world and ensuring they benefit from advanced AI technology and new revenue models.”

OpenAI’s partnership with Axel Springer comes on the heels of concerns from creators, authors, and publishers who have criticized and sued generative AI companies for training their models on their content without consent or compensation. Some publishers like The New York Times, Vox Media, BBC News, Reuters, and CNN have blocked OpenAI from accessing their data. Striking deals with AI companies, however, could provide a brand new revenue source for publishers who are currently going through the worst year for the media business in decades.

As part of the deal, Alex Springer will provide OpenAI with both current news articles as well as archives from all its brands to train its large language models, the foundational tech that powers ChatGPT. When ChatGPT uses Axel Springer’s articles in its responses, it will include attribution and links to the pieces for transparency. Axel Springer will also be able to use OpenAI’s technology to improve its own products, The Wall Street Journal reported.

This isn’t the first deal that OpenAI has struck with a news publisher. Earlier this year, the company entered into a two-year partnership with The Associated Press to use select content from the AP’s archives dating back to 1985 to train its AI models, although the terms of that deal do not include letting ChatGPT use AP content in its responses. OpenAI also has a $5 million partnership with the American Journalism Project to explore how local news organizations can benefit from artificial intelligence.

Apple will reward record labels and artists who offer their music in spatial audio, a relatively new audio format that is more immersive than regular stereo. According to a Bloomberg report, artists who release their music in spatial audio will receive “added weighting” starting next year. That, Bloomberg speculates, could mean higher royalties.

Nearly all of Apple’s audio hardware such as AirPods, HomePod, the iPhone and the upcoming Vision Pro headset, support playback in spatial audio, so the company’s move to incentivize artists is almost certainly to ensure that most music available on Apple Music is available in a format that Apple has positioned as a selling point. Notably, Bloomberg notes that Apple Music listeners wouldn’t necessarily have to stream a song in spatial audio for artists to be rewarded. Simply having their music available in the format would be enough.

Apple added spatial audio, which is powered by technology from Dolby Atmos, to Apple Music in 2021. Most of the company’s original shows and movies on Apple TV+ are also offered in the format. Artists also have the option of mixing their older music in the new format, something that bands from all decades are already doing. Mixing music in the format isn’t wildly expensive, according to Bloomberg; if true, this push could help get independent musicians and smaller acts on board.

Some Apple Music competitors like Amazon Music and Tidal also offer spatial audio on their services. But Spotify, Apple’s biggest music stream rival, is a notable exception, even though rumors about a high-quality music format on the service have swirled for years.

Genetic testing company 23andMe changed its terms of service to prevent customers from filing class action lawsuits or participating in a jury trial days after reports revealing that attackers accessed personal information of nearly 7 million people — half of the company’s user base — in an October hack.

In an email sent to customers earlier this week viewed by Engadget, the company announced that it had made updates to the “Dispute Resolution and Arbitration section” of its terms “to include procedures that will encourage a prompt resolution of any disputes and to streamline arbitration proceedings where multiple similar claims are filed.” Clicking through leads customers to the newest version of the company’s terms of service that essentially disallow customers from filing class action lawsuits, something that more people are likely to do now that the scale of the hack is clearer.

“To the fullest extent allowed by applicable law, you and we agree that each party may bring disputes against the other party only in an individual capacity and not as a class action or collective action or class arbitration,” the updated terms say. Notably, 23andMe will automatically opt customers into the new terms unless they specifically inform the company that they disagree by sending an email within 30 days of receiving the firm’s notice. Unless they do that, they “will be deemed to have agreed to the new terms,” the company’s email tells customers.

23andMe did not respond to a request for comment from Engadget.

In October, the San Francisco-based genetic testing company headed by Anne Wojcicki announced that hackers had accessed sensitive user information including photos, full names, geographical location, information related to ancestry trees, and even names of related family members. The company said that no genetic material or DNA records were exposed. Days after that attack, the hackers put up profiles of hundreds of thousands of Ashkenazi Jews and Chinese people for sale on the internet. But until last week, it wasn’t clear how many people were impacted.

In a filing with the Securities and Exchange Commission, 23andMe said that “multiple class action claims” have already been against the company in both federal and state court in California and state court in Illinois, as well as in Canadian courts.

Forbidding people from filing class action lawsuit, as Axios notes, hides information about the proceedings from the public since affected parties typically attempt to resolve disputes with arbitrators in private. Experts, such as Chicago-Kent College of Law professor Nancy Kim, an online contractor expert, told Axios that changing its terms wouldn’t be enough to protect 23andMe in court.

The company’s new terms are sparking outrage online. “Wow they first screw up and then they try to screw their users by being shady,” a user who goes by Daniel Arroyo posted on X. “Seems like they’re really trying to cover their asses,” wrote another user called Paul Duke, “and head off lawsuits after announcing hackers got personal data about customers.”

Google is bringing Gemini, the new large language model it just introduced, to Android, beginning with the Pixel 8 Pro. The company’s flagship smartphone will run Gemini Nano, a version of the model built specifically to run locally on smaller devices, Google announced in a blog post. The Pixel 8 Pro is powered by the Google Tensor G3 chip designed to speed up AI performance.

This lets the Pixel 8 Pro add several smarts to existing features. The phone’s Recorder app, for instance, has a Summarize feature that currently needs a network connection to give you a summary of recorded conversations, interviews, and presentations. But thanks to Gemini Nano, the phone will now be able to provide a summary without needing a connection at all.

Gemini smarts will also power Gboard’s Smart Reply feature. Gboard will suggest high-quality responses to messages and be aware of context in conversations. The feature is currently available as a developer preview and needs to be enabled in settings. However, it only works with WhatsApp currently and will come to more apps next year.

“Gemini Nano running on Pixel 8 Pro offers several advantages by design, helping prevent sensitive data from leaving the phone, as well as offering the ability to use features without a network connection,” wrote Brian Rakowski, Google Pixel’s vice president of product management.

As part of today’s AI push, Google is upgrading Bard, the company’s ChatGPT rival, with Gemini as well, so you should see significant improvements when using the Pixel’s Assistant with Bard experience. Google is also rolling out a handful of AI-powered productivity and customization updates on other Pixel devices, including the Pixel Tablet and the Pixel Watch, although it isn’t immediately clear what they are.

Google

Gemini Nano is the smallest version of Google's large language model, while Gemini Pro is a larger model that will power not just Bard but other Google services like Search, Ads and Chrome, among others. Gemini Ultra, Google's beefiest model, will arrive in 2024 and will be used to further AI development.

Although today’s updates are focused on the Pixel 8 Pro, Google spoke today about AI Core, an Android 14 service that allows developers to access AI features like Nano. Google says AI Core is designed run on “new ML hardware like the latest Google Tensor TPU and NPUs in flagship Qualcomm Technologies, Samsung S.LSI and MediaTek silicon.” The company adds that “additional devices and silicon partners will be announced in the coming months.”

Rockstar has released the trailer of Grand Theft Auto VI, the next game in the blockbuster video game franchise a day earlier than expected. You can watch it here:

It's been a decade since Rockstar Games released Grand Theft Auto V. While fans have been more than able to keep themselves busy with GTA Online and a few re-releases, they've been waiting patiently (or impatiently) for more single-player action. The wait's now much shorter as Rockstar has revealed the first official look at the game's debut trailer.

As indicated by a recent teaser image and various rumors, GTA VI will be set in Vice City, Rockstar's take on Miami. Given the weapons and other tech on display, it seems to be a contemporary game, rather than one set in the '80s like 2002's Grand Theft Auto: Vice City. It also appears that, as rumored, the game will have a playable female character for the first time in the franchise.

A major leak last year saw 90 videos of a GTA VI test build emerge online. The clips indicated that the game would mark a return to Vice City and that there'd be two playable characters.

GTA VI will be released in 2025, 12 years after its predecessor, which is the second-best selling video game of all time.

This is a developing story and will be updated with more details.

Last week, a team of researchers published a paper showing that it was able to get ChatGPT to inadvertently reveal bits of data including people’s phone numbers, email addresses and dates of birth that it had been trained on by asking it to repeat words “forever”. Doing this now is a violation of ChatGPT’s terms of service, according to a report in 404 Media and Engadget’s own testing.

“This content may violate our content policy or terms of use”, ChatGPT responded to Engadget’s prompt to repeat the word “hello” forever. “If you believe this to be in error, please submit your feedback — your input will aid our research in this area.”

There’s no language in OpenAI’s content policy, however, that prohibits users from asking the service to repeat words forever, something that 404 Media notes. Under “Terms of Use”, OpenAI states that users may not “use any automated or programmatic method to extract data or output from the Services” — but simply prompting the ChatGPT to repeat word forever is not automation or programmatic. OpenAI did not respond to a request for comment from Engadget.

The chatbot’s behavior has pulled back the curtain on the training data that modern AI services are powered by. Critics have accused companies like OpenAI of using enormous amounts of data available on the internet to build proprietary products like ChatGPT without consent from people who own this data and without compensating them.

A team of researchers was able to make ChatGPT reveal some of the bits of data it has been trained on by using a simple prompt: asking the chatbot to repeat random words forever. In response, ChatGPT churned out people’s private information including email addresses and phone numbers, snippets from research papers and news articles, Wikipedia pages, and more.

The researchers, who work at Google DeepMind, the University of Washington, Cornell, Carnegie Mellon University, the University of California Berkeley, and ETH Zurich, urged AI companies to seek out internal and external testing before releasing large language models, the foundational tech that powers modern AI services like chatbots and image-generators. “It’s wild to us that our attack works and should’ve, would’ve, could’ve been found earlier,” they wrote, and published their findings in a paper on Tuesday that 404 Media first reported on.

Chatbots like ChatGPT and prompt-based image generators like DALL-E are powered by large language models, deep learning algorithms that are trained on enormous amounts of data that critics say is often scraped off the public internet without consent. But until now, it wasn’t clear what data OpenAI’s chatbot was trained on since the large language models that power it are closed-source.

When the researchers asked ChatGPT to “repeat the word ‘poem’ forever”, the chatbot initially compiled, but then revealed an email address and a cellphone number for a real founder and CEO”, the paper revealed. When asked to repeat the word “company”, the chatbot eventually spat out the email address and phone number of a random law firm in the US. “In total, 16.9 percent of the generations we tested contained memorized [personally identifiable information]” the researchers wrote.

Using similar prompts, the researchers were also able to make ChatGPT reveal chunks of poetry, Bitcoin addresses, fax numbers, names, birthdays, social media handles, explicit content from dating websites, snippets from copyrighted research papers and verbatim text from news websites like CNN. Overall, they spent $200 to generate 10,000 examples of personally identifiable information and other data cribbed straight from the web totalling “several megabytes”. But a more serious adversary, they noted, could potentially get a lot more by spending more money. “The actual attack”, they wrote, “is kind of silly.”

OpenAI patched the vulnerability on August 30, the researchers say. But in our own tests, Engadget was able to replicate some of the paper’s findings. When we asked ChatGPT to repeat the word “reply” forever, for instance, the chatbot did so, before eventually revealing someone’s name and Skype ID. OpenAI did not respond to Engadget’s request for comment.

Sonos is planning to enter new markets over the next couple of years. The company, which currently makes smart speakers and soundbars, is gearing up to launch high-end, over-the-ear headphones, smaller in-ear wireless earbuds and a streaming box for TVs, according to a Bloomberg report. The new headphones would directly compete with Apple’s AirPods Max and AirPods, as well as devices from Sony and Bose, while the streaming box would compete with the Apple TV and devices from Roku and Amazon.

Sonos reportedly plans to charge more than $400 for the over-the-ear headphones — around the same as Sony's $400 highly-regarded WH-1000MX5, but cheaper than Apple' $549 AirPods Max. Its streaming box will reportedly cost between $150 and $200 — significantly higher than Apple, Amazon and Roku’s devices.

The company, best known for high-end smart speakers, has been trying to develop headphones since 2019. Its new products are an attempt to revive sales, which have slumped this year. On an earnings call earlier this month, Sonos CEO Patrick Spence said that he expects brand new products to account for a “large portion” of the company’s revenue by the second half of 2024.

Sonos plans to release the over-the-ear headphones by April next year. Bloomberg reports that they will come in black and white color options, sync with other Sonos speakers, and also let people use their voice to navigate between songs. Sonos is also working on software to let people control all Sonos equipment in their house from their phones when they are away.

The streaming box will reportedly be released by the end of 2024 or early 2025. It will run an Android-based operating system and run streaming apps, similar to how existing streaming boxes work. Sonos is currently in talks with Netflix and companies offering live television to support its device.

In addition to headphones and a streaming box, Sonos is also working on updated versions of current home theater equipment, including a new subwoofer, a second-generation Roam speaker, a new version of its high-end Arc soundbar, and a more expensive Era 100 speaker with an ethernet port that Bloomberg says is aimed at businesses. The updated Arc might cost $,1200, a major price bump compared to the current model that retails for $900. It isn't clear how much the Era 100, the Roam and the subwoofer will cost. 

Spotify struck a special deal with Google that lets it pay no commission to Google when people sign up for subscriptions using the music streaming service’s own payment system on Android, according to new testimony in the ongoing Epic v. Google trial first reported by The Verge. As part of the same deal, Spotify paid Google just four percent commission if users signed up for the service through Google, far less than most other apps which typically pay 15 percent for subscriptions through the Google Play Store.

“Listening to music is one of [the phone’s] core purposes… if we don’t have Spotify working properly across Play services and core services, people will not buy Android phones”, Google’s partnerships head Don Harrison reportedly said in court. Both Google and Spotify also agreed to put $50 million each in a “success fund” as part of the deal.

The remarks were made as part of a lawsuit first filed against Google by Epic Games, the maker of the wildly popular Fortnite, in 2020. Epic claimed that Google’s Play Store on Android was an illegal monopoly that forced app makers to part with huge sums of cash in exchange for offering users in-app purchases through the Play Store. Epic filed a similar lawsuit against Apple in 2021, which it lost.

Spotify initially supported Epic in its fight against Google and Apple. But in 2022, the company started using a Google program called User Choice Billing that let Android apps use their own payment systems in exchange for giving a reduced cut to Google. The special deal revealed in court showed that Google was willing to carve out even more exceptions for popular apps like Spotify.

Google has had some pretty big business secrets spilled in the last few days. Last week, an economics professor testifying on behalf of the company in a separate antitrust trial that has since wrapped up, revealed that Google pays Apple 36 percent of all ad revenue it generates through Apple’s Safari browser, a figure which Alphabet CEO Sundar Pichai later confirmed while he was testifying in the Epic v. Google trial.

The Verge also reported earlier this month that Google offered Netflix, another popular streaming service, a custom deal. It offered a reduced commission of 10 percent, which Netflix turned down – instead choosing to not offer users a way to sign up for Netflix directly within its Android app.