Twitter is beginning to roll out its long-promised encrypted direct messaging feature. However, the initial rollout comes with some major limitations that could make it less than ideal for privacy-conscious Twitter users.

Of note, the feature is currently only available to verified Twitter users, which includes Twitter Blue subscribers and those part of a “Verified Organization.” It’s not clear if this is just for the early rollout or if encryption will be added to the growing list of exclusive features for users with a checkmark. For now, an encrypted chat requires both users to be verified, according to the company.

There are also some significant limitations to the feature itself. It doesn’t support group messages, or any kind of media other than links. The company also doesn’t allow users to report an encrypted message directly, advising on a help page that users should report accounts separately if they “encounter an issue with an encrypted conversation participant.”

Twitter

Finally, the level of encryption appears to be less secure than what other apps offer. For one, message metadata is not encrypted. Furthermore, Twitter notes that “currently, we do not offer protections against man-in-the-middle attacks” and suggests that the company itself is still able to access encrypted DMs without the participants knowing. “If someone–for example, a malicious insider, or Twitter itself as a result of a compulsory legal process—were to compromise an encrypted conversation, neither the sender or receiver would know,” the company explains on a help page. It added that it’s working on improvements that would make such exploits more “difficult.”

That’s particularly notable because it falls far short of the standard Twitter owner Elon Musk has described when expressing his desire to add encryption for Twitter DMs. He has said he wants it to be impossible for the company to access users’ encrypted messages even if “someone puts a gun to our heads.”

In a tweet, Twitter security engineer Christopher Stanley acknowledged the shortcoming. “We’re not quite there yet, but we’re working on it.”

For those who are verified and want to try out the feature anyway, encrypted messaging can be accessed via the info menu (that’s the same menu you use to block or report a conversation) within a particular DM. Once encryption is enabled, the encrypted messages will appear as a separate message thread with labels at the top of the chat to indicate that the conversation is encrypted.

Google and WhatsApp have confirmed they are aware of a bug that makes it appear as if WhatsApp is accessing phones’ microphones unnecessarily on some Android devices. The issue first cropped up a month ago, but gained new attention after a Twitter engineer tweeted about it in a post that was boosted by Elon Musk.

An image shared by Twitter engineer Foad Dabiri appeared to show that the microphone had been repeatedly running in the background while he wasn’t using the app. He tweeted a screenshot from Android’s Privacy Dashboard, which tracks how often apps access a device’s microphone and camera.

Musk retweeted Dabiri’s post, saying “WhatsApp cannot be trusted.” Incidentally, Musk is known to be a fan of Signal, and has said encrypted direct messages on Twitter could roll out as soon as this month. The company didn’t respond to a request for comment.

In a statement shared on Twitter, WhatsApp suggested it was an Android-related issue, and not a result of inappropriate microphone access by the messaging app “We believe this is a bug on Android that mis-attributes information in their Privacy Dashboard and have asked Google to investigate and remediate,” the company said.

Dabiri is not the first to notice the issue. WhatsApp blogwabetainfohighlighted the bug a month ago, describing it at the time as “a false positive” affecting owners of some Pixel and Samsung devices. They added that restarting the phone may be a possible fix. Meanwhile, Google has said little about what could be causing the discrepancy, but confirmed it’s looking into the matter. "We are aware of the issue and are working closely with WhatsApp to investigate,” a Google spokesperson said in a statement.

Meta is bringing more ads to Reels on Facebook and Instagram, and changing up how creators can earn money from their content. The change comes two months after Meta “paused” a bonus program that rewarded creators for hitting specific goals.

Now, Meta says it’s “accelerating” its plan to bring ads into Reels on Facebook, with “thousands” of creators now eligible to join the monetization program. The company plans to test a similar program for Reels creators on Instagram in the “coming weeks.”

Notably, Meta is structuring the program a bit differently than typical revenue sharing arrangements. The company says it will pay creators using “a new payout model that pays creators based on the performance of their public reels, not the earnings of ads on their reels” and that “initially, payouts in the test will be determined by the number of plays.” In other words, creators who rack up the highest number of views will earn the most, regardless of other engagement metrics or how much ad revenue their clips generate.

The decision to prioritize views apparently came as a result of the company’s earlier tests, with Meta noting that payments based on ad performance can sometimes negatively impact creators for things out of their control, like the company’s own ad inventory. Those terms are also likely a reflection of the fact that while Reels drive a lot of engagement for Meta, the feature hasn’t proven to be as easily monetized as its feeds and stories features. (The company says in its blog post that it’s continuing to “optimize the ad experience for advertisers,” on Reels.)

It’s also not yet clear just how much creators can expect to make through the new program. Meta had previously promised Reels creators monthly bonuses up to $35,000 a month when it launched the Reels Play bonus program in 2021. But the company reportedly slashed those payments last year, before “pausing” the program entirely in march. According to a Meta spokesperson, the ‘overarching goal” is for creators to be able to earn consistent payouts, but these numbers “will vary widely by creator.”

Meta is bringing more ads to Reels on Facebook and Instagram, and changing up how creators can earn money from their content. The change comes two months after Meta “paused” a bonus program that rewarded creators for hitting specific goals.

Now, Meta says it’s “accelerating” its plan to bring ads into Reels on Facebook, with “thousands” of creators now eligible to join the monetization program. The company plans to test a similar program for Reels creators on Instagram in the “coming weeks.”

Notably, Meta is structuring the program a bit differently than typical revenue sharing arrangements. The company says it will pay creators using “a new payout model that pays creators based on the performance of their public reels, not the earnings of ads on their reels” and that “initially, payouts in the test will be determined by the number of plays.” In other words, creators who rack up the highest number of views will earn the most, regardless of other engagement metrics or how much ad revenue their clips generate.

The decision to prioritize views apparently came as a result of the company’s earlier tests, with Meta noting that payments based on ad performance can sometimes negatively impact creators for things out of their control, like the company’s own ad inventory. Those terms are also likely a reflection of the fact that while Reels drive a lot of engagement for Meta, the feature hasn’t proven to be as easily monetized as its feeds and stories features. (The company says in its blog post that it’s continuing to “optimize the ad experience for advertisers,” on Reels.)

It’s also not yet clear just how much creators can expect to make through the new program. Meta had previously promised Reels creators monthly bonuses up to $35,000 a month when it launched the Reels Play bonus program in 2021. But the company reportedly slashed those payments last year, before “pausing” the program entirely in march. According to a Meta spokesperson, the ‘overarching goal” is for creators to be able to earn consistent payouts, but these numbers “will vary widely by creator.”

Bluesky, the Jack Dorsey-backed decentralized social network, isn’t publicly available yet but it’s already become one of the buzziest Twitter alternatives. Scoring an invite to the service, which has been in a closed beta for barely two months, has become a sought-after status symbol with invite codes selling for hundreds of dollars on eBay.

The iOS app has been downloaded more than 375,000 times, according to analytics from data.ai, and the app has already become one of the most popular social media apps in the Play Store barely two weeks after launching. Those numbers may not seem particularly impressive, but considering the app only has about 65,000 users – it’s clear that demand for the service is far outstripping the number of available invites.

What is Bluesky?

Bluesky started in 2019 as an internal project at Twitter with the goal of creating an open source decentralized standard for social media. Dorsey, a longtime proponent of decentralized networks, had long championed the idea of such a standard, saying he hoped Twitter could one day “be a client of this standard.”

Bluesky went independent in 2021, and officially ended its association with Twitter in 2022 following Musk’s takeover of the company. Bluesky is now overseen by CEO Jay Graber, along with a small team of developers. Dorsey is still attached to the project — he has a seat on Bluesky’s board — but isn’t running the show like he once did at Twitter. “Jack doesn’t have unilateral power,” Graber wrote in a recent post on Bluesky. “I have the most control over this endeavor.”

Dorsey, who these days posts more often on Nostr, his other decentralized social media of choice, is at least an occasional Bluesky user. He recently took to Bluesky to share his latest thoughts on Musk and how he has handled the Twitter acquisition. The former CEO, who once said “Elon is the singular solution I trust,” said he thought Musk should have walked away from the Twitter deal, and that the company “went south” after his acquisition.

The service itself looks and feels a lot like Twitter. Bluesky defaults to a chronological timeline of accounts you follow, but there’s also an algorithmic “what’s hot” feed that surfaces popular posts or “skeets” as early users have dubbed them. It doesn’t yet have as much functionality as Twitter — for example, there are no direct messages or video support — but the interface is almost identical to Twitter.

Bluesky is also invitation-only, which has so far helped fuel its viral success. The platform began opening its waitlist more widely last month, and new users receive one invite code every two weeks (though some of the more prolific users get codes more often). Graber has said the invite system is meant to prevent spammers and others from easily gaining access to the platform, and to ensure it grows “organically.”

How is it different from Mastodon?

Both Bluesky and Mastodon are open source and decentralized, but there are some significant differences between the two. Mastodon is built on a protocol called Activity Pub, which has since been embraced by other platforms like Medium, Flipboard and even Meta.

Bluesky is building its own protocol, called the AT Protocol or Authenticated Transfer Protocol. As the company notes on its website, Bluesky, the app, is meant to show off what the AT Protocol can do, similar to how early browsers demonstrated the potential of the web.

According to Bluesky, one of the biggest differences between AT and other decentralized social networking standards is how it handles account portability. “With the AT Protocol, you can move your account from one provider to another without losing any of your data or social graph,” the company explains.

Bluesky

Right now, Bluesky itself is running the only instance of the platform. But just as Mastodon has thousands of servers available, there could eventually be numerous Bluesky servers run by other entities. That’s also where the emphasis on account portability comes in: it should make moving between servers fairly seamless, including taking your social graph with you when you move, which isn’t currently possible on Mastodon.

Bluesky is also taking an interesting approach to algorithms and moderation. On Mastodon, each server sets its own norms and moderation guidelines, which are enforced by admins. But while Bluesky’s team is to some extent setting their own norms for what’s currently acceptable, the app allows users to customize what level of moderation they want, ranging from “hide,” “warn,” or “show,” for things like hate speech, spam and impersonation.

Bluesky

Right now, there are some moderation labels that appear on posts containing things like impersonation, but Graber says the goal is for labels to eventually be community-led so that moderations can function independently of individual servers. Graber has dubbed the approach “composable moderation.”

Similarly, the platform is planning a customizable take on algorithms. Graber has said Bluesky wants to create a “marketplace of algorithms” so users can be in control of how their feeds are filtered.

Why is it suddenly so popular?

Like many other Twitter-like services, there’s been increased interest in Bluesky since Elon Musk took over the company. Its early ties to Dorsey and Twitter has also fueled the curiosity of those searching for a new platform.

But it wasn’t until Bluesky began opening more of its waitlist over the last couple weeks that it started to become more widely known. As the service approached its first 50,0000 users last week, more and more prominent Twitter users have joined. Representative Alexandria Ocasio Cortez, James Gunn, Chrissy Teigen, Rian Johnson and other celebrities have joined in recent days. So have other recognizable Twitter users like WeRateDogs, Dril, darth and Musk nemesis ElonJet.

It’s not just former blue checks though. Black and transgender voices have also found Bluesky welcoming. So have sex workers and shitposters. Needless to say, things have also gotten very weird, very quickly.

The massive influx of users exposed a technical glitch that soon became known as the “hellthread.” The glitch meant that long threads with lots of replies would break in the app’s main feed but users would continue to see notifications for new replies. “Just don’t reply to the thread to stay out,” Graber posted. Bluesky users, naturally, had other ideas. People not only kept replying, they tagged others into the phantom thread. At some point, users began dropping nudes into the hellthread, ultimately causing the app’s developers to block nudity from Bluesky’s “What’s Hot” feed.

Weirdness and nudity aside, Bluesky’s current popularity seems to be driven by both novelty of the service and because it’s so much smaller and simpler than its predecessor. The small user base and invite system has also so far mostly kept out trolls and bots and spam that tend to make Twitter a lot less usable for high-profile accounts or people from marginalized backgrounds. At a time when Twitter has reversed rules protecting trans people, unbanned Nazis and harassers, and algorithmically favors those paying for blue checks, Bluesky is a reminder of how good “old Twitter” used to be,

Will it last?

Just how long the good vibes will last is another question. Bluesky isn’t the first Twitter alternative to take off and see a surge in interest from former Twitter power users. Mastodon surged for months last year following Musk’s takeover of Twitter, and inspired high-profile Twitter users and developers to flock to the platform. But that growth seems to have now leveled off, and the number of active users has declined after months of growth.

Bluesky has some advantages over Mastodon: it’s easier to use and much more intuitive for Twitter users, assuming you can score an invite. But whether Bluesky remains as popular as it is now will likely depend on if the platform can sustain the same level of interest once it expands beyond the current invite system. Many viral apps start off as invite only (Clubhouse, for example) only to fade into the background once they are no longer exclusive and novel.

On the other hand, Twitter is still a mess and there is clearly sustained demand for a non-Musk operated alternative.

As the buzz around ChatGPT and other generative AI increases, so has scammers’ interest in the tech. In a new report published by Meta, the company says it’s seen a sharp uptick in malware disguised as ChatGPT and similar AI software.

In a statement, the company said that since March of 2023 alone, its researchers have discovered “ten malware families using ChatGPT and other similar themes to compromise accounts across the internet” and that it’s blocked more than 1,000 malicious links from its platform. According to Meta, the scams often involve mobile apps or browser extensions posing as ChatGPT tools. And while in some cases the tools do offer some ChatGPT functionality, their real purpose is to steal their users’ account credentials.

In a call with reporters, Meta Chief Security Officer Guy Rosen said the scammers behind these exploits are taking advantage of the surge in interest in Generative AI. “As an industry we've seen this across other topics that are popular in their time such as crypto scams fueled by the immense interest in digital currency,” Rosen said. “So from a bad actor’s perspective, ChatGPT is the new crypto.”

Meta noted that people who manage businesses on Facebook or who otherwise use the platform for work have been particular targets. Scammers will often go after users’ personal accounts in order to gain access to a connected business page or advertising account, which are more likely to have a linked credit card.

To combat this, Meta said it plans to introduce a new type of account for businesses called “Meta Work” accounts. These accounts will enable users to access Facebook’s Business Manager tools without a personal Facebook account. “This will help keep business accounts more secure in cases when attackers begin with a personal account compromise,” the company said in a statement. Meta said it will start a “limited” test of the new work accounts this year and will expand it “over time.”

Additionally, Meta is rolling out a new tool that will help businesses detect and remove malware. The tool “guides people step-by-step through how to identify and remove malware, including using third-party antivirus tools” to help prevent businesses from repeatedly losing access to accounts.

Meta’s researchers aren’t the first to warn about fake ChatGPT tools leading to hacked accounts. Recently, researchers warned about a Chrome extension posing as ChatGPT software that led to the hacking of a number of Facebook accounts. The exploit, reported byBleeping Computer, became known as the “Lily Collins” hack because the names on victims’ accounts were changed to “Lilly Collins.”

During a call with reporters, Meta’s Head of Security Policy, Nathaniel Gliecher, said these attacks also often target people connected to businesses. “What they'll want to do is to close that personal account to burn their access and prevent the legitimate user from getting back in,” he said. “One of the tactics we're now seeing is where they will take the personal account and rename it to have the name of a prominent celebrity in hopes that that gets the account taken down.” He added that the new Work Accounts would help prevent similar hacks in the future.

Snap is bringing ads to two if its most buzzy features: its TikTok clone, Spotlight, and its OpenAI-powered chatbot “My AI.” The company announced the new ad formats during its presentation at the NewFronts advertising event.

Snap began testing ads in Spotlight last year, but will now begin to roll them out globally. The addition could be a significant boost to Snap’s advertising business, which has taken a hit in recent months. Spotlight, which rewards creators for popular short-form videos, has more than 350 million monthly users according to the company.

The addition of ads could also open up some interesting possibilities for creators on Spotlight – mainly, the potential for revenue sharing arrangements. Though Snap hasn’t disclosed any such plans for now, the company did recently expand other ad revenue sharing programs, so it’s not hard to imagine the company might one day consider such a move in Spotlight as well.

Separately, Snap is also bringing ads to My AI, the chatbot it recently rolled out to all users. There, My AI will be able to surface sponsored links directly in chats with users. The company says the sponsored links are meant to be relevant to the conversation. For example, the chatbot, which can already provide recommendations for local restaurants based on the app’s Snap Map, may surface a sponsored post from a local restaurant or food delivery service in response to a question about where to eat dinner. Likewise, a conversation about travel or video games could bring up sponsored links from an airline or game retailer.

The company describes the My AI ads as an early experiment, but the fact Snap is choosing to bring ads to the chatbot so soon after its launch is notable. The company says it’s seen high engagement with the feature, which unlike friend-to-friend chats doesn’t keep conversations private. But the company’s recent decision to put the bot at the top of everyone’s chats has also been controversial, so introducing ads at this stage could also potentially turn off some users. Regardless, the experiment will be an interesting test for how social media companies can potentially monetize generative AI features.

LinkedIn is experimenting with a new generative AI feature for job hunters. The company is testing a new feature that will generate brief, cover letter-like messages candidates can send to hiring managers on the platform. The feature is starting to roll out now for the site's premium subscribers.

With the update, users will see the option to “Let AI draft a message to the hiring team” alongside open roles on the platform’s jobs page. The feature draws on “information from your profile, the hiring manager’s profile, the job description, and the company of interest” to create a “highly personalized” message, according to the company.

LinkedIn

In the example provided by LinkedIn, the message reads like the opening few sentences of a cover letter. However, the AI-written message users see will likely vary based on how much information is in your LinkedIn profile. The company notes that “customization is still important,” and that users should double check and edit the text before sending it.

LinkedIn, which is owned by OpenAI partner Microsoft, has experimented with other generative AI features. The platform added AI writing suggestions to profiles, and “collaborative articles” which also make use of AI-written text.

Nextdoor is introducing its first generative AI feature, an in-app “assistant” that can help users rewrite “potentially unkind” posts on the neighborhood social network. The new feature is rolling out “over the next several weeks.”

It’s far from the first time the company has experimented with ways to remind users to keep conversations “neighborly.” The company, which has at time struggled to fight the perception that its platform can be toxic, began using “kindness reminders” in 2019 and last year introduced pop-ups reminding users to be more “empathetic.” The app has also served up more targeted nudges to promote anti-racist language and less heated political discussions.

The new “assistant” takes the approach a step further. Instead of preemptive reminders, the assistant will prompt users to “rephrase potentially unkind comments” and suggest new wording for the post. “It's great to tell people, ‘hey, be a little bit more constructive,’ or ‘you don't always have to respond,’” Nextdoor CEO Sarah Friar says. “But now you're actually helping them reframe it in a way they might not have thought of.”

Friar notes that the assistant-written posts are optional and that users are free to make changes to the text it suggests. “It’s saying ‘when you word things this way,’ or when you maybe just add a few more sentences or add some context or an emoji, you can help someone hear you,” Friar says. “They may not agree with you, but at least they can begin to hear you.”

The assistant itself relies on the same OpenAI models used by ChatGPT, and has also been trained on “all of the data we've collected over our almost 10 years of life,” according to Friar. She says this allows the assistant to provide more tailored suggestions for other kinds of posts on the platform. In an example provided by the company, the assistant rewrites a post from a user looking for landscaping work with new phrasing the app says may “get a better response” than the original.

While the assistant is Nextdoor’s first use of generative AI, it likely won’t be the last. Friar says she’s particularly interested in how generative AI can be used for recommendations to help small businesses using the platform.

Clubhouse, the once hyped audio app that took off during the height of the pandemic, is laying off staff as part of a company “reset,” the startup announced. In a letter to staff shared by the company, founders Paul Davison and Rohan Seth said they were “scaling back our org by over 50% and saying goodbye to many talented, dedicated teammates.”

Interestingly, Davison and Seth said that the company, which was reportedly valued at $4 billion in 2021, was still well-funded, with “years of runway remaining.” They added that they “do not feel immediate pressure to reduce costs.”

Clubhouse exploded in popularity early in 2021 at a time when many people were feeling socially isolated due to the pandemic. The app was able to draw in big names like Elon Musk and Oprah for its live audio chats, which helped further fuel its early success. But Davison would later acknowledge that growth came “way too fast,” and usage dropped off amid competitors cloning the service and more people returning to their normal activities as the pandemic eased.

In their letter to staff, Davison and Seth alluded to these issues. “But as the world has opened up post-Covid, it’s become harder for many people to find their friends on Clubhouse and to fit long conversations into their daily lives,” they wrote. “To find its role in the world, the product needs to evolve. In order to fix this we need to reset the company, eliminate roles and take it down to a smaller, product-focused team.”

The founders went on to say that they will be focusing on “Clubhouse 2.0,” though they didn’t elaborate on what the service may look like going forward, just that it would be built by a “smaller team.”

Clubhouse isn’t the only live audio service to struggle. Spotify recently shut down its standalone live audio app, Greenroom. And Reddit closed its take on live audio, Reddit Talk, in March. Even Amazon and Meta have struggled to keep some of their audio projects going.