We got a healthy dose of Star Wars and Marvel shows on Disney Plus this year, but the more mature series from Hulu helped balance things out.

Read the full story at The Verge.

Asus is planning to launch a new Rog Strix laptop at CES on January 6th, 2025, the company confirmed in a post spotted by VideoCardz. The short teaser shared by Asus shows a laptop with RGB lighting that wraps all the way around the bottom of the device, likely making for an even more colorful underglow when compared to previous generations.

Though Asus doesn’t say which Rog Strix models it will introduce, leaked retail listings suggest Asus could reveal new Rog Strix 18 Scar and Rog Strix G16 laptops.

While the Rog Strix Scar 18 is rumored to come with up to an Intel Core Ultra 9 285 HX processor and Nvidia GeForce RTX 5090 graphics chip, the Rog Strix G16 could feature the same CPU options but with up to a GeForce RTX 5080, as noted by NotebookCheck.

Asus may have more to share than just a pair of new Rog Strix laptops at CES, which is just a couple of weeks away. Recent leaks also indicate that Asus is getting ready to reveal an upgraded Rog Flow Z13 hybrid tablet / gaming laptop equipped with AMD’s next-gen “Strix Halo” processor.

This year’s Prime Video streaming content was led by adaptations and spinoffs like Fallout and The Lord of the Rings: The Rings of Power.

Read the full story at The Verge.

Earlier this year, TCL released a trailer for Next Stop Paris — an AI-animated short film that seems like a Lifetime movie on steroids. The trailer had all the hallmarks of AI: characters that don’t move their mouths when they talk, lifeless expressions, and weird animation that makes it look like scenes are constantly vibrating.

I thought this might be the extent of TCL’s experimentation with AI films, given the healthy dose of criticism it received online. But boy, was I wrong. TCL debuted five new AI-generated short films that are also destined for its TCLtv Plus free streaming platform, and after the Next Stop Paris debacle, I just had to see what else it cooked up.

Though the new films do look a little better than Next Stop Paris, they serve as yet another reminder that AI-generated videos aren’t quite there yet, something we’ve seen with many of the video generation tools cropping up, like OpenAI’s Sora. But in TCL’s case, it’s not just the AI that makes these films bad.

Here are all five of them, ranked from tolerable (5) to “I wish I could unsee this” (1).

5. Sun Day

This futuristic short film basically has the same concept as Ray Bradbury’s short story “All Summer in...

Read the full story at The Verge.

A federal jury in Delaware determined on Friday that Qualcomm didn’t breach its agreement with Arm through its 2021 acquisition of Nuvia, a startup founded by three former Apple engineers. As reported earlier by Bloomberg and Reuters, the decision stems from a two-year-long legal battle that accused Qualcomm of misusing the chip designs Arm licensed to Nuvia before its acquisition.

Despite delivering a win for Qualcomm, the jury couldn’t determine whether Nuvia breached its agreement with Arm, meaning the case can be tried again. “I don’t think either side had a clear victory or would have had a clear victory if this case is tried again,” US District Court Judge Maryellen Noreika said, according to Reuters.

Qualcomm bought Nuvia for $1.4 billion to bolster the company’s lineup of next-generation chips, like the Snapdragon X chips inside current Copilot Plus laptops. Still, testimony during the trial revealed that Qualcomm's internal documents also showed the company projected it could save as much as $1.4 billion every year on payments to Arm.

In 2022, Arm ignited a legal battle after Qualcomm continued to pay its existing royalty fees to Arm, which were allegedly much lower than what Nuvia was paying. After the two failed to come to an agreement, Arm argued the designs licensed to Nuvia were no longer valid, and that Qualcomm should destroy the technology created with them.

During an interview on Decoder this week, Arm CEO Rene Haas couldn’t share much about the trial, but said, “The principles as to why we filed the claim are unchanged.”

The jury ultimately sided with Qualcomm after viewing Arm’s internal documents that estimate Arm could’ve lost $50 million in revenue as a result of Nuvia’s acquisition, according to Reuters. This week, Nuvia co-founder Gerard Williams also testified that the startup only used “one percent or less” of Arm technology in its finished technology, Reuters reported.

“The jury has vindicated Qualcomm’s right to innovate and affirmed that all the Qualcomm products at issue in the case are protected by Qualcomm’s contract with ARM,” Ann Chaplin, Qualcomm’s general counsel and corporate secretary, said in an emailed statement to The Verge. “We will continue to develop performance-leading, world class products that benefit consumers worldwide, with our incredible Oryon ARM-compliant custom CPUs.”

The Verge reached out to Arm with a request for comment but didn’t immediately hear back.

The US government has charged a dual Russian and Israeli national with allegedly building and maintaining LockBit’s malware code, while receiving over $230,000 in cryptocurrency for his work. The 51-year-old Rostislav Panev was arrested in Israel pending extradition to the US, making him the third member of the LockBit ransomware group in custody.

Authorities previously arrested other alleged members of the LockBit group, including Mikhail Vasiliev and Ruslan Magomedovich Astamirov, both of whom have pleaded guilty to various charges, including conspiracy to commit computer fraud.

Authorities are still searching for Lockbit’s alleged ringleader, Dmitry Khoroshev, with a reward worth up to $10 million. The DOJ claimed in May that “Khoroshev alone allegedly received at least $100 million in disbursements of digital currency through his developer shares of LockBit ransom payments,” based on a 20 percent share of ransom payments extorted by affiliates who used the group’s software.

As outlined in the complaint, Panev is accused of working as a developer for LockBit since the group first formed in 2019, helping to wage ransomware attacks on hundreds of entities around the globe, including hospitals, businesses, government agencies, and more.

Law enforcement linked Panev to LockBit after finding login credentials on his computer for a dark web repository housing “multiple versions of the LockBit builder,” which is the tool that allowed members “to generate custom builds of the LockBit ransomware malware for particular victims.”

Panev allegedly admitted to writing and maintaining LockBit’s malware code in interviews with the Israeli police. Some of the code he’s said to have created can disable Windows Defender antivirus software, run malware on multiple computers on a network, and print LockBit’s ransom note on all the printers in a victim’s network. Panev claimed he didn’t realize he was involved in illegal activity at first, according to the complaint.

YouTube is taking a tougher stance on clickbait, saying it will remove content with titles or thumbnails that promise viewers “something that the video doesn’t deliver,” as spotted earlier by TechCrunch. This change will “slowly” roll out in India first, according to YouTube’s blog post, but will “expand to more countries” in the “coming months,” YouTube spokesperson Jack Malon says in a statement to The Verge.

YouTube says the policy will combat “egregious” clickbait that misleads viewers, with a particular focus on videos related to “breaking news” or “current events.” The company’s examples of egregious clickbait include a video with the title “the president resigned!” that doesn’t actually address a resignation or a “top political news” thumbnail attached to a video with no news content.

As the policy rolls out in India, YouTube will remove content that violates the rules without giving a strike to creators, at least at first. “And as we continue to educate creators, our enforcement efforts will prioritize new video uploads moving forward,” YouTube says.

The Consumer Financial Protection Bureau (CFPB) has filed a lawsuit against Zelle and three banks that own it — Wells Fargo, Bank of America, and JPMorgan Chase — claiming they failed “to protect consumers from widespread fraud.” Zelle is a payment network designed to compete with payment platforms like Venmo and Cash App, but the CFPB says the banks “rushed” it to market, enabling fraud that’s cost consumers more than $870 million since it launched in 2017.

The lawsuit cites Zelle’s designs and features, including a “limited” identity verification process that involves assigning a “token” to a user’s email address or mobile phone number that they can use to verify their account with a one-time passcode. This setup makes it easier for scammers to take over accounts, as well as hide their own identities or pretend to be other institutions, the CFPB alleges.

CFPB complaint

One of the most common Zelle scams involves bad actors impersonating a financial institution or a federal agency, who then trick customers into sending them money. After facing pressure from the CFPB, the banks backing Zelle started issuing refunds to victims of this type of scam last year. This latest lawsuit follows other CFPB actions to tighten regulation around digital wallet apps and payment networks.

The CFPB accuses Zelle and the banking trio of failing to track and quickly stop criminals on the platform, as they allegedly didn’t relay information about known fraudulent transactions with other institutions in the payment network. It also alleges Bank of America, JPMorgan Chase, and Wells Fargo didn’t properly address the risk of fraud despite the “hundreds of thousands” of complaints they received.

Zelle pushed back on the lawsuit in a statement published on Friday. “The CFPB’s attacks on Zelle are legally and factually flawed, and the timing of this lawsuit appears to be driven by political factors unrelated to Zelle,” Zelle spokesperson Jane Khodos said. “The CFPB’s misguided attacks will embolden criminals, cost consumers more in fees, stifle small businesses and make it harder for thousands of community banks and credit unions to compete.”

The CFPB is asking the court to stop Zelle’s parent company, Early Warning Services, and the banks from violating consumer protection laws, and compensate users, among other penalties.

Google Fiber is changing up its internet plans in Huntsville, Alabama and Nashville, Tennessee. The new Core 1 Gig, Home 3 Gig, and Edge 8 Gig plans appear to have launched last month and streamline the company’s existing options, as spotted earlier by 9to5Google.

These options replace the 1 Gig, 2 Gig, 5 Gig, and 8 Gig plans currently available in other supported cities. Like the existing 1 Gig plan, Google Fiber’s new Core 1 Gig option costs $70 / month with symmetrical 1 gigabit per second upload and download speeds. It comes with GFiber’s Multi-Gig Wi-Fi 6E Router and supports up to one mesh extender.

Screenshot: Google

The $100 / month Home 3 Gig plan sits between the $100 / month 2 Gig and $125 / month 5 Gig options, offering up to 3-gig speeds, a GFiber Multi-Gig Wi-Fi 6E router, up to two mesh extenders, along with priority room optimization that brings “additional wired connectivity to the rooms that matter most.” Subscribers can also add an internet battery backup for an extra $10 per month, which offers up to two hours of “full-bandwidth uptime, with no internet slow downs” in case a power outage knocks out your router or fiber jack.

Lastly, Google’s $150 / month 8 Gig Edge “always-on” plan comes with upload and download speeds of up to 8 gigabits per second, the same GFiber Wi-Fi 6E router, and up to two mesh extenders. It also offers up to 5,000 square feet of coverage, priority room optimization, an included internet backup battery, and a 25 percent refund if your internet goes down for over 45 minutes.

When asked whether Google Fiber will bring these new plans to more locations, Google spokesperson Sunny Gettinger said the company will have “more to share in the new year.” Since Google Fiber’s inception in 2010, the company has been gradually expanding its fiber footprint across the US while achieving faster speeds. Google rolled out a 20-gig Wi-Fi 7 plan in select cities last year.

Update, December 20th: Added a response from Google.

Google is planning to add a new “AI Mode” to its search engine, according to a report from The Information. The company will reportedly display an option to switch to AI Mode from the top of the results page, allowing you to access an interface similar to its Gemini AI chatbot.

The new AI Mode tab would live on the left side of the “All,” “Images,” “Videos,” and “Shopping” tabs, The Information reports. When you receive a response in AI Mode, The Information says Google will display links to related webpages and “a search bar below the conversational answer that prompts users to ‘Ask a follow-up...’”

This tracks with Android Authority’s report from earlier this month, which spotted an AI Mode in a beta version of the Google app. 9to5Google also dug up code suggesting you can use AI Mode to ask questions using your voice. The Verge reached out to Google with a request for comment but didn’t immediately hear back.

With OpenAI rolling out search in ChatGPT for all users, Google is likely under increased pressure to consolidate search and AI. The company already displays AI search summaries for some queries and recently expanded the feature to dozens of more countries in October.

A Google Street View image played in role in solving a missing person investigation in Spain, according to reports from the BBC and The New York Times. The image, which can still be seen on Street View, shows what appears to be a person loading a large object into the trunk of a car, as Gizmodo points out. Another image showed someone “transporting a large white bundle in a wheelbarrow,” the BBC reports.

Spanish National Police:

Part of the human remains of the missing person have been found buried in a cemetery in a town in the province of Soria and advanced technical means had to be used to locate them.

One of the clues that investigators had to solve the crime, although it was not decisive, was some images that they detected during the investigations in a location search application.

Screenshot: Google

Though the press release doesn’t explicitly name the location application it used, a spokesperson for the Spanish National Police confirmed to The New York Times that it was Google Maps while also saying, “The image was not the key to solving the case.”

Authorities were investigating the case of a 33-year-old man who disappeared last year while visiting a partner in Tajueco, Soria, according to the Spanish outlet El País. Police arrested a woman and her ex-partner in connection with the crime in November.

Google Maps has been credited with helping investigators in the past, with Italian authorities using Street View to trace the location of a mafia fugitive in 2022.

Google has introduced a new AI “reasoning” model capable of answering complex questions while also providing a rundown of its “thoughts,” as reported earlier by TechCrunch. The model, called Gemini 2.0 Flash Thinking, is still experimental and will likely compete with OpenAI’s o1 reasoning model.

In a post on X, Google DeepMind chief scientist Jeff Dean says the model is “trained to use thoughts to strengthen its reasoning,” and also benefits from the speed that comes along with the faster Gemini Flash 2.0 model. The demo shared by Dean shows how Gemini 2.0 Flash Thinking goes about answering a physics problem by “thinking” through a series of steps before offering a solution.

This isn’t necessarily “reasoning” in the way humans perform it, but it means the machine breaks down instructions into smaller tasks that can produce stronger outcomes.

Another example, posted by Google product lead Logan Kilpatrick, shows the model reasoning its way through a problem that involves both visual and textual elements. “This is just the first step in our reasoning journey,” Kilpatrick says. You can try out Gemini 2.0 Flash Thinking on Google’s AI Studio.

There have been quite a few notable updates in the AI space as of late, with Google revealing its upgraded Gemini 2.0 model earlier this month as part of the company’s push into “agentic” AI. Meanwhile, OpenAI made the full version of its o1 reasoning model available to ChatGPT subscribers.

The Federal Aviation Administration banned drones from flying over “critical New Jersey infrastructure” in the wake of thousands of drone sightings for which there seem to be very few satisfying answers. The ban will last until January 17th and will block unauthorized drone activity in 22 areas due to “special security reasons,” according to ABC7NY.

Under the FAA’s temporary flight restriction, drones can’t operate within one nautical mile, from the ground up to an altitude of 400 feet, in the affected airspace. A Notice to Airmen (NOTAM) on the FAA’s website says the government “may use deadly force” against unmanned aircraft in restricted areas if it “poses an imminent security threat,” while the pilot may be “intercepted, detained and interviewed” by law enforcement.

During a background call last week, an FAA official said the agency issued its first temporary flight restrictions over Morris County, New Jersey, and the Trump National Golf Club in Bedminster in late November. This latest ban covers a much wider swath of the state, including major hubs like Jersey City, Elizabeth, Bayonne, Harrison, and Camden.

For weeks, millions of people in the New York City metro area have traded concerns and conspiracy theories about the mysterious drones zipping through the skies. Despite sightings popping up even outside of New Jersey, government officials have little information to share on the situation.

A joint statement from the FAA, FBI, Department of Homeland Security, and the Department of Defense maintains that the drones don’t appear to “present a national security or public safety risk,” while attributing some sightings to “lawful commercial drones, hobbyist drones, and law enforcement drones, as well as manned fixed-wing aircraft, helicopters, and stars mistakenly reported as drones.”

Still, lawmakers are pushing for answers — and possibly even legislation that will make it easier to track and monitor drone activity.

Microsoft is previewing live translation on Intel and AMD-based Copilot Plus PCs. The feature is rolling out now to Windows 11 Insiders in the Dev Channel, allowing users to translate audio from over 44 languages into English subtitles.

Live translation, which initially launched on Qualcomm-powered Copilot Plus PCs, works with any audio played through a Copilot Plus PC, whether it’s coming from a YouTube video, a live video conference, or a recording. If the audio is in a supported language, Windows 11 will display real-time captions in English. The feature can currently translate from Spanish, French, Russian, Chinese, Korean, Arabic, and more.

Image: Microsoft

Microsoft has been gradually bringing more AI features to Intel and AMD-powered Copilot Plus PCs. Earlier this month, Microsoft began testing Recall, which takes snapshots of your activity on a Copilot Plus PC and lets you call up specific memories, on devices with Intel and AMD chips.

Microsoft is also rolling out an update to live translation on Qualcomm-equipped Copilot Plus PCs, as Windows 11 Insiders in the Dev Channel can now translate select languages to Simplified Chinese.

Intel’s new Arrow Lake CPUs came out in October promising high performance at lower power levels, but reviews, including ours, noted surprisingly lackluster gaming performance. Tom Warren said of the Core Ultra 9 285K, “...in many titles, it provides worse performance than the 14th Gen chips it was designed to replace.”

Now, the troubled chipmaker says it has been gradually releasing updates that fixed most of the issues it identified, as reported earlier by Tom’s Hardware.

“At the end of the day, there were four root issues that we needed to address, and fixes for those are actually already in the field right now,” Robert Hallock, Intel’s VP and GM of client AI and technical marketing said during an interview with HotHardware. “They’ve been coming out over the last two weeks or so in drips and drabs, as update schedules allow.”

One issue stemmed from a “mistimed” update that was supposed to optimize Windows’ processor power management (PPM) system for Intel’s Core Ultra 200S-series processors. (PPM adjusts the performance of a CPU based on the current power plan, such as Balanced, High Performance, Power Saver, etc).

Instead of releasing the PPM update before reviewers got their hands on the processors, Intel says it scheduled the update to go out when it became widely available, which may have caused reviewers to see worse-than-expected performance. Due to the missing PPM update, Intel’s performance-boosting Application Performance Optimizer (APO) also couldn’t take effect in games, while “misconfigured” performance settings also had a negative impact on reviewers’ benchmarks.

Image: Intel

A fix was already applied for these issues in Windows 11 build 26100.2161. Intel also says Epic Games resolved a driver compatibility issue causing the blue screen of death when running games with Easy Anti-Cheat, such as Star Wars Outlaws.

Intel is planning to release another set of performance upgrades for Arrow Lake CPUs in January, and it will “provide a comprehensive performance update” at CES. For now, Intel recommends updating Windows and applying the latest BIOS update to your motherboard if you haven’t already. Otherwise, you can wait until the “final” performance update next year.

Threads is rolling out a new feature that lets you share someone’s photos and videos without including the original post. You’ll be able to add your own text alongside the reshared media, which will include a credit to the creator, according to Instagram head Adam Mosseri.

“This is a quick, easy way to add your creative takes to trending images and clips without quote posting,” Mosseri says. The original poster’s username will appear in the top left corner of the photo or video, with a repost counter in the lower left. Threads spokesperson Alec Booker tells The Verge that tapping on the username will show a list of posts using the media, with the original post appearing at the top.

Besides making it a little cleaner to reshare media, the feature is likely designed to help cut down on complaints about users reposting content without credit. Creators will receive a notification when their posts are reused, and will also have the ability to turn resharing off from their account settings.

To repost an image or video on Threads, you can long-press the media (or the repost button) and select use media. From there, you can add your own text and share the post. X has a similar option, though only for videos, which includes a credit beneath the video and a link to the original post.

Update, December 18th: Added more information from Threads.

Roku got the exclusive US streaming rights to next year’s X Games Aspen and a “soon-to-be-announced” X Games event in the summer, the company announced on Wednesday. Both events will air on Roku’s free 24/7 sports channel as the service continues its push into live sports.

Though the X Games were founded by ESPN in the ‘90s, the sports network sold majority ownership to MSP Sports in 2022. Since then, the X Games have streamed on various platforms, including YouTube, Twitch, ESPN, ABC, and the VR app Xtadium.

But now, X Games Aspen will appear on the Roku Sports Channel, a newly launched hub for Roku’s live Sunday MLB games, Formula E races, and other sports-related content. This channel lives within the overarching Roku Channel that comes pre-installed on most Roku devices and is also available on the web, as well as an app on third-party smart TVs and mobile phones.

Additionally, Roku launched a free ad-supported streaming TV channel dedicated to the X Games, which will air “programming highlights, clips, interviews, archival content, and more” leading up to the event. X Games Aspen takes place from January 23rd to the 25th, with more details to come on an “additional” X Games coming this summer.

Roku is just one of many streaming platforms getting into sports, with Netflix airing live NFL games on Christmas Day, Max launching a live sports add-on, and Amazon Prime Video picking up streaming rights to NBA games. Even the tournament series Street League Skateboarding signed a deal with the right-wing streaming platform Rumble.

It’s been more than two decades since Seagate began working on heat-assisted magnetic recording (HAMR) — and now the company may finally be ready to release a hard drive using the technology. A new product page spotted by Tom’s Hardware shows an Exos M hard drive sporting up to 32TB of storage using Seagate’s Mozaic 3 Plus HAMR platform.

Seagate’s Mozaic 3 Plus technology allows for bigger hard drive capacities by making data bits smaller and closer together on each disk. To write data, a laser diode attached to the drive’s recording heads heats small areas of the disk. “Each bit is heated and cools down in a nanosecond, so the HAMR laser has no impact at all on drive temperature, or on the temperature, stability, or reliability of the media overall,” Seagate writes on its website.

Seagate says its Exos M hard drive has a 3TB per platter density, making it useful for enterprise applications like powering AI systems. We still don’t know when Seagate could release its Exos M hard drive, as its product page currently shows a link to “Stay Informed,” but a launch seems imminent.

As pointed out by Tom’s Guide, Seagate said in a filing earlier this month that it had “successfully completed qualification testing” for its HAMR hard drives with “several customers within the Mass Capacity markets, including a leading cloud service provider.” It says it will start shipping its HAMR-based hard drive to the unnamed cloud provider in the “coming weeks.”

The Verge reached out to Seagate with a request for more information but didn’t immediately hear back.

Seagate isn’t the only company working on high-capacity hard drives. In October, Western Digital launched a 32TB hard drive using energy-assisted perpendicular magnetic recording (ePMR), while Toshiba recently demonstrated high-capacity hard drives with HAMR and microwave-assisted magnetic recording (MAMR).

Grubhub has agreed to pay $25 million to settle a Federal Trade Commission lawsuit that claimed the food delivery service misled customers and drivers while also damaging the reputation of restaurants. The proposed settlement will require Grubhub to make several changes to the platform, such as showing the total delivery cost when customers place an order.

Along with advertising “highly inflated hourly pay rates for drivers,” the FTC’s initial complaint accused Grubhub of hiding “the true cost of its services” by adding delivery fees that raised the price of customers’ final orders. The agency claimed that starting around 2019, Grubhub began advertising lower delivery fees to attract more customers but then began tacking on a “service” fee that increased the cost of orders anyway.

The FTC also alleged the company charged Grubhub Plus members for delivery despite advertising the subscription as having “free” or “$0” deliveries. The agency claimed Grubhub makes the plan easy to sign up for but difficult to cancel while also allegedly blocking the accounts of users with large gift card balances.

Screenshot: FTC

Additionally, the FTC accused Grubhub of adding restaurants to the platform even if they never signed up to sell food on the service. “Grubhub has had as many as 325,000 unaffiliated restaurants on its platform — more than half of all of the available restaurants on Grubhub,” the FTC claims. As a result, many customers wound up having issues with their orders, resulting in bad feedback for unaffiliated restaurants.

Grubhub is now required to show customers the full cost of delivery and can no longer add “junk fees” to orders. It’s also banned from listing unaffiliated restaurants on the platform, and can only make driver earnings claims “that it can back up with evidence and in writing.” Grubhub must also notify customers when they’re banned and offer a way to appeal the decision, as well as make it easier to cancel Grubhub Plus.

“While we categorically deny the allegations made by the FTC, many of which are wrong, misleading or no longer applicable to our business, we believe settling this matter is in the best interest of Grubhub and allows us to move forward,” Grubhub spokesperson Najy Kamal said in a statement to The Verge. The company also responded to the settlement in a post on its website.

Though Grubhub was initially ordered to pay $140 million, it is “partially suspended based on the company’s inability to pay the full amount.” The company's $25 million will go toward refunding affected customers, but the FTC says the full judgment will be due “immediately” if Grubhub “is found to have misrepresented its financial status.”

YouTube is partnering with the Creative Artists Agency (CAA) to help creators identify content using their AI-generated likenesses on the platform and submit removal requests. The company will test the controls with celebrities and athletes early next year before rolling it out to “top YouTube creators, creative professionals, and other leading partners representing talent.”

In September, YouTube announced plans for tools that would help manage AI-generated depictions of creators and their voices. Now, the company says it can give celebrities (and soon, creators) the ability to manage AI copies of their likeness, such as their face, “at scale.”

Last year, CAA introduced the CAAVault, which scans and stores the digital likenesses of its clients, including their faces, bodies, and voices.

YouTube is also working on “synthetic-singing identification technology” that will detect AI content that attempts to replicate creators’ singing voices. YouTube has already started letting music labels request the removal of AI content that simulates an artist’s voice, and also began requiring creators to label videos containing AI-generated content earlier this year.