Neuralink will be able to surgically implant its device into another patient’s brain. The Wall Street Journal reports that the company was approved to move forward with a second procedure months after Noland Arbaugh became the first person to receive the brain implant.

Elon Musk said last week that the company was “accepting applications for the second participant” in the trial. The company began recruiting potential participants for its first clinical trial last year with the goal of bringing the technology to people with ALS, spinal cord injuries or other conditions that cause quadriplegia.

Neuralink has also reportedly come up with a potential fix for an issue that caused Arbaugh’s implant to malfunction about a month after his surgery. The company said earlier this month that some of the implant’s threads “retracted from the brain” causing the issue. Arbaugh recently told Bloomberg that software updates have since restored many of those capabilities. Neuralink has shared clips of Arbaugh, who is paralyzed from the neck down, playing chess, controlling a music player app and performing other activities. 

According to The Journal, Neuralink told the FDA that in a second procedure it would place the implant’s threads deeper into the patient’s brain to prevent them from moving as much as they did in Arbaugh’s case. The FDA is apparently on board with the changes. The company reportedly wants to complete the second surgery in June and has seen more than 1,000 people sign up for a chance to participate in the trial.

Meta’s advertising policies are once again in the spotlight as a watchdog group says the company approved more than a dozen “highly inflammatory” ads that broke its rules. The ads targeted Indian audiences and contained disinformation, calls for violence and conspiracy theories about the upcoming elections.

The ads are detailed in a new report from Ekō, a nonprofit watchdog organization. The group says it submitted the ads as a “stress test” of Meta’s company’s advertising systems, but that the spots “were created based upon real hate speech and disinformation prevalent in India.”

In all, the group was able to get 14 of 22 ads approved through Meta’s company’s advertising tools even though all of them should have been rejected for breaking the company’s rules. The group didn’t disclose the exact wording of the ads, but said they “called for violent uprisings targeting Muslim minorities, disseminated blatant disinformation exploiting communal or religious conspiracy theories prevalent in India's political landscape, and incited violence through Hindu supremacist narratives.” Researchers at Ekō pulled the ads before they ran and they were never seen by actual Facebook users, according to the report.

It’s not the first time Ekō has gotten inflammatory ads approved by Meta in an effort to draw attention to its advertising systems. The group previously got a batch of hate-filled Facebook ads targeting users in Europe approved, though the ads never ran.

In its latest report, Ekō says it also used generative AI tools to create images for the ads. Researchers at the organizations said none of the ads were flagged by Meta as containing AI-generated material, despite the company’s statements that it’s working on systems to detect such content.

Meta didn’t immediately respond to a request for comment. In a response to Ekō, the company pointed to its rules requiring political advertisers to disclose their use of AI and a blog post about its efforts to prepare for the Indian elections.

Microsoft Paint is getting new image generation powers with a new tool called Cocreator. Powered by "diffusion-based algorithms," Cocreator can generate images based on text prompts as well as your own doodles in the Paint app.

The company has been experimenting with AI image generation in Paint for a while, and early versions of Cocreator have been available to developers and Windows Insiders since the fall. But with the introduction of CoPilot+ PCs, the feature is now official.

During a demo at its Surface event, the company showed off how Cocreator combines your own drawings with text prompts to create an image. There’s also a “creativity slider” that allows you to control how much you want AI to take over compared with your original art. As Microsoft pointed out, the combination of text prompts and your own brush strokes enables faster edits. It could also help provide a more precise rendering than what you’d be able to achieve with DALL-E or another text-to-image generator alone.

Catch up on all the news from Microsoft's Copilot AI and Surface event today!

The web version of Threads could soon be much more useful. Meta is starting to test custom Tweetdeck-like feeds that will allow users to track multiple topics, searches and accounts in a single view.

People who are part of the test can set “pinned columns” that will track updates around specific topics, tags, accounts or search terms. Users can also opt to have these columns automatically refresh with new content. Based on screenshots shared by Mark Zuckerberg, the new Threads columns look a lot like Tweetdeck, the desktop app long favored by Twitter’s power users. The app is now called X Pro and only available to X’s paid subscribers.

The test is the latest sign Meta is looking to make Threads a more reliable source for real-time information. The company has also added a “recent” tab and trending topics to search. But being able to track multiple feeds of updates at once is even more useful. It could also address long-running complaints about Threads’ algorithmic “for you” feed, which tends to surface a random mix of days-old posts and bizarre personal stories from unconnected accounts.

It’s not clear how many people will be part of Meta’s initial test of the feature, though Adam Mosseri said the company is looking for feedback on the changes. But the company has often rolled out major Threads changes to small group of users first before making them more widely available.

Meta’s Oversight Board has accepted its first case involving a post on Threads and it will allow the group to weigh in on the debate over the role of political content on Threads. The board, which started taking appeals from Threads users earlier this year, announced its first case involving Meta’s newest app.

The case stems from a post by a Japanese user who was replying to a screenshot of a news article about Prime Minister Fumio Kishida and allegations of tax evasion. The reply, according to the board, included “several hashtags using the phrase ‘drop dead.’” Meta’s content moderators removed the post, citing the company’s rules against inciting violence. But after the user appealed to the Oversight Board and had the case accepted, Meta reversed course, saying that the post didn’t violate its rules after all.

All that may sound like a fairly typical case for the board, which regularly reviews Meta’s content moderation decisions and pushes the social media company to change its policies. But it’s the first time the group will apply that same process to Threads. And the board has suggested it will use the case to weigh in on the company’s controversial decision to stop showing political content in its algorithmic recommendations on Threads and Instagram.

“The Board selected thi case to examine Meta’s content moderation policies and enforcement practices on political content on Threads,” the Oversight Board wrote in a statement. “This is particularly important, in the context of Meta’s decision not to proactively recommend political content on Threads.”

As usual, it will likely be several months before we see the Oversight Board’s decision actually play out in any policy changes at Meta. In the meantime, the board is seeking public comment on “how Meta’s choice not to recommend political content on Threads and Instagram newsfeeds, or pages not followed by users, affects access to information and political speech.”

Threads is inching closer to becoming an actually useful source for real-time news and updates. The app is finally rolling out the ability to search posts in order of recency, after testing the feature last month.

“In an effort to make it easier to find timely, relevant content on Threads, we’re introducing a Recent tab for your searches,” Instagram’s Adam Mosseri wrote in an update. “Search results here are still evaluated for quality, but you can now see them in chronological order.”

The change has been a long requested one from users hoping Meta’s app will one day be a source of breaking news and real-time information the way that Twitter historically functioned. Being able to search for topics and keywords and find the most recent results is key to finding up-to-date details and commentary about breaking news, sports and anything else happening in real time.

On the other hand, Meta has also made it clear that it would prefer “news” to not be what Threads is known for. Mosseri has said he doesn’t want to “encourage” hard news on Threads and the company actively discourages political content. Threads’ default “for you” algorithm is also known for surfacing days-old posts, random personal stories and other content that’s not exactly timely.

It’s also worth pointing out that Threads’ new recency filter in search is not the same as the “latest” search filter on X. As Mosseri noted in his post, Meta still hides an unknown number of posts in search results that have been “evaluated for quality,” so Threads search will never surface all of the posts containing your search terms. But being able to at least find posts that aren’t a few days old should make looking for timely information a lot less frustrating.

A group of TikTok creators have joined the legal fight to keep the app from being banned in the United States. Eight creators have sued the US government in an effort to block a law requiring TikTok's parent company ByteDance to sell the service. 

The lawsuit claims that the “Protecting Americans from Foreign Adversary Controlled Applications Act” is unconstitutional because it violates the First Amendment rights of the creators who depend on the platform. “They have found their voices, amassed significant audiences, made new friends, and encountered new and different ways of thinking—all because of TikTok’s novel way of hosting, curating, and disseminating speech,” it states. “The Act’s ban of TikTok threatens to deprive them, and the rest of the country, of this distinctive means of expression and communication.”

The lawsuit comes one week after TikTok filed its own lawsuit against the government. According to The Washington Post, the company is “covering” the legal fees for the creators participating in the latest suit. It’s also strategy that has worked for the company in the past. A group of Montana-based TikTok creators sued the state over an attempted statewide ban last year. That effort was ultimately successful and the ban never went into effect. The Montana creators were represented by the same law firm currently repping the eight creators involved in the latest suit.

As Google starts to make its latest video-generation tools available, the company says it has a plan to ensure transparency around the origins of its increasingly realistic AI-generated clips. All video made by the company’s new Veo model in the VideoFX app will have digital watermarks thanks to Google’s SynthID system.

SynthID is Google’s digital watermarking system that started rolling out to AI-generated images last year. The tech embeds imperceptible watermarks into AI-made content so that AI detection tools can recognize that the content was generated by AI. Considering that Veo, the company’s latest video generation model previewed onstage at I/O, can create longer and higher-res clips than what was previously possible, tracking the source of such content will be increasingly important.

During a briefing with reporters, DeepMind CEO Demis Hassabis said that SynthID watermarks would also expand to AI-generated text. As generative AI models advance, more companies have turned to watermarking amid fears that AI could fuel a new wave of misinformation. Watermarking systems would give platforms like Google a framework for detecting AI-generated content that may otherwise be impossible to distinguish. TikTok and Meta have also recently announced plans to support similar detection tools on their platforms and label more AI content in their apps.

Of course, there are still significant questions about whether digital watermarks on their own offer sufficient protection against deceptive AI content. Researchers have shown that watermarks can be easy to evade. But making AI-made content detectable in some way is an important first step toward transparency.

Catch up on all the news from Google I/O 2024 right here!

It’s no secret that Meta is eager to prove itself as an AI company, and it might be eyeing a new hardware project to help it get there. The company is in the early stages of “exploring” designs for AI-enabled headphones, according to a new report in The Information.

The company is reportedly calling it “Camerabuds,” because the headphones or earbuds (the company is considering both designs) would have two outward-facing cameras that would be able to detect the wearer’s surroundings and power real-time AI features. In that way, it sounds a but like what the company has done with the multimodal AI features in its smart glasses.

For now, it’s unclear if such a project would actually move forward and Meta didn’t immediately respond to a request for comment. Mark Zuckerberg has reportedly “seen several possible designs for the device” but hasn’t been “satisfied” with any so far. Internally there are also engineering concerns, like issues around battery life and heat, as well as the myriad of privacy issues that come with camera-enabled wearables. People with long hair could also pose a potential challenge as their locks would block the cameras’ view.

As The Information points out, Meta has a somewhat mixed track record when it comes to hardware projects. It killed its lineup of smart speakers, called Portal, in 2022 and axed a long-rumored camera-enabled smartwatch. But the most recent version of its Ray-Ban-branded smart glasses, which feature Meta AI, have been more positively received. The company hasn’t disclosed how many pairs it’s sold, but Zuckerberg has said that there's been “high demand” and that the glasses have seen better and more sustained engagement than the first-generation shades.

Putting Meta AI into another piece of hardware would also be very much in line with the company’s ambitions to compete with other leading AI firms. While AI-specific hardware like Rabbit’s R1 and Humane’s AI Pin have so far been underwhelming, Meta is likely hoping that putting its latest AI features into a device people are already used to wearing, like headphones, will be an easier sell. 

Google is teasing an intriguing new AI feature one day ahead of its I/O developer conference. The company shared a brief video on X that appears to show a new camera-powered AI feature that’s able to recognize what’s in the frame in real time.

The video, which is labeled as a “prototype,” shows what appears to be a Pixel device with the camera open viewing the keynote stage at I/O. The person holding the camera asks, “hey, what do you think is happening here?”

A voice replies that “it looks like people are setting up for a large event, perhaps a conference or presentation.” It’s also able to identify the “IO” letters as being tied to Google’s developer conference and mentions “new advancements in artificial intelligence.” As the two voices go back and forth, a text transcript appears on the screen.

It’s not exactly clear what the feature is, though it bears some similarities to Google Lens, the company’s camera-powered search feature. What’s shown in the teaser video, however, appears to be working in real-time, and responding to voice commands much like the multimodal AI in Meta’s smart glasses. The fact that the demo is shown on a Pixel device is also intriguing as Google often releases new AI-powered features on its Pixel lineup first.

While it’s somewhat unusual for Google to preview one of its announcements so soon before its big keynote, it’s likely not a coincidence that the company dropped the video right as OpenAI showed off similar capabilities with its new GPT-4o model during a live event. Whatever Google has in store, though, we don’t have much longer to wait to get the full details. Google I/O kicks off tomorrow, May 14, and Engadget will be covering the keynote live from Mountain View.