Techin' Geek

Meta on Wednesday announced it’s expanding ads on Threads to all users globally. The expansion, which starts next week, will be gradual, the company says, noting it may take months for the full rollout to complete.

Meta CEO Mark Zuckerberg has repeatedly praised the social app and X rival, which has now grown to over 400 million monthly active users, as the company’s next big hit. Shortly after the app’s debut, the exec told investors that Threads had a good chance of reaching 1 billion users in a few years’ time.

The app has rapidly grown since its July 2023 debut, reaching 200 million users by mid-2024, 320 million as of January 2025, and then adding another 30 million as of last April, before reaching its latest milestone.

Though Threads hasn’t yet hit that milestone, the company has been testing ads for some time. A year ago, Threads began testing ads in the U.S. and Japan, and last April, the platform opened up to global advertisers.

The company has made it easy for existing advertisers to expand their reach to include Threads by allowing them to automatically place ads through both Meta’s Advantage+ program and via manual campaigns. Supported formats include image and video, ads, including the newer additions of the 4:5 aspect ratio format and carousel ads. Advertisers can manage their Threads ads alongside those for Facebook, Instagram, and WhatsApp in their Business Settings, which simplifies cross-posting.

In addition, Meta expanded the third-party verification already available on Facebook and Instagram (including feed and Reels) to the Threads feed through Meta Business Partners. This provides advertisers with independent brand safety and suitability verification for Threads — a useful feature when Threads’ nearest rival is dealing with a scourge of illegal deepfakes.

The company didn’t share how often users would see ads in their feed, but noted that ad delivery initially would remain “low” as the feature scaled to global users.

Days ahead of a scheduled trial, social media company Snap has settled a lawsuit accusing the platform of causing social media addiction, according to reports from multiple outlets.

According to the New York Times, the settlement was announced Tuesday in the California Superior Court in Los Angeles County. The lawsuit against Snap was brought by a 19-year-old known in court documents as K.G.M., accusing the social media app of designing algorithms and features that caused addiction and mental health issues.

The terms of the settlement were not disclosed.

The lawsuit also names other platforms, including Meta, YouTube, and TikTok. No settlement has been reached with these platforms. Notably, Snap is still a defendant in other similar social media addiction cases filed against it.

According to documents unveiled in the ongoing cases, Snap employees raised concerns around risks to the mental health of teens dating back at least nine years. The company has said these examples were “cherry-picked” and were taken out of context.

Plaintiffs in these cases are drawing parallels to Big Tobacco — referring to lawsuits in the 1990s against cigarette companies that concealed health risks — alleging that the platforms obscured information about potential harms from their users. They argue that features like infinite scroll, auto video play and algorithmic recommendations have tricked users into continuously using apps, leading to depression, eating disorders, and self-harm, according to NYT.

Snap CEO Evan Spiegel had been scheduled to testify in the trial, which would have marked the first time a social media company faces a jury in an addiction lawsuit — no platform has lost such a case at trial yet. The remaining case against Meta, TikTok, and YouTube is set to proceed with jury selection beginning next Monday, January 27, with Meta CEO Mark Zuckerberg expected to take the witness stand.

If plaintiffs prevail, legal experts predict the cases could result in multibillion-dollar settlements and potentially force platforms to redesign their products. But the companies have so far defended themselves in part by arguing that those same design choices — like algorithmic recommendations, push notifications, and infinite scroll — are similar to a newspaper deciding what stories to publish and are protected speech under the First Amendment.

Snap did not immediately respond to a request for comment.

Industry reports and the growth of voice model companies in the Indian market suggest that there is a growing demand for voice AI solutions in the country. Voice is a popular medium for communication among people and businesses in India. That’s why enterprises and startups are eager to use voice AI to be more efficient at customer support, sales, customer acquisition, hiring, and training.

But recognizing market demand is one thing — proving businesses will pay is another. Y Combinator rejected the application from Bolna, a voice orchestration startup built by Maitreya Wagh and Prateek Sachan, five times before finally accepting it into the fall 2025 batch, skeptical that the founders could turn interest into revenue.

“When we were applying for Y Combinator, the feedback we got was, ‘great to see that you have a product that can create realistic voice agents, but Indian enterprises are not going to pay, and you are not going to make money out of this,’” Wagh told TechCrunch.

The startup applied with the same idea for the fall batch but was able to show it had revenue of more than $25,000 coming in every month for the last few months. At that time, the company was running $100 pilots to help users build voice agents. Now the startup is pricing those pilots at $500.

The momentum has continued. The startup said on Tuesday that it has raised a $6.3 million seed round led by General Catalyst, with participation from Y Combinator, Blume Ventures, Orange Collective, Pioneer Fund, Transpose Capital, and Eight Capital. The round also includes individual investors, including Aarthi Ramamurthy, Arpan Sheth, Sriwatsan Krishnan, Ravi Iyer, and Taro Fukuyama.

The product and customers

Bolna is building an orchestration layer — essentially a platform that connects and manages different AI voice technologies — akin to startups like Vapi, LiveKit, and VoiceRun, to suit the idiosyncrasies of interactions in India, including noise cancellation, getting verification on the caller ID platform Truecaller, and handling mixed languages.

Feature-wise, the company has built specific nuances for Indian users, such as speaking numbers in English regardless of the core language, or allowing for keypad input for longer inputs.

Wagh noted that the key differentiation of Bolna is that it makes it easy for users to build voice agents by just describing them, even if they don’t know much about the underlying technology, and start using them for calls. The company said that 75% of its revenue is coming from self-serve customers.

He also said that because Bolna is an orchestration layer, it doesn’t depend on a single model, so enterprises can easily switch when there is a better model available.

“Our platform allows customers to switch models easily or even use different models for different locales to get the best out of them. An orchestration layer is necessary for enterprises to ensure they are getting the best models because one model can be better today and another one can be better tomorrow,” Wagh said.

The company has a range of clients, including car reselling platform Spinny, on-demand house-help startup Snabbit, beverage companies, and dating apps. Most of these are small to midsize businesses that use Bolna’s self-serve platform.

Separately, Bolna is pursuing large enterprise deals. For these large enterprises and custom implementations, Bolna has a team of forward-deployed engineers — specialists who work directly with clients on-site or closely with their teams. The startup has signed two large enterprises as paying customers and has four more in the pilot stage. Currently, Bolna employs nine forward-deployed engineers and is adding two to three people to that team every month to support this enterprise push.

Bolna has seen steady growth in both call volumes and revenue. It say it’s now handling over 200,000 calls per day and on the verge of crossing $700,000 in annual recurring revenue (ARR). The company noted that while 60% to 70% of call volume is in English or Hindi, other regional languages are steadily rising.

Akarsh Shrivastava, who is part of the investment team at General Catalyst, said that the firm found Bolna impressive because its orchestration layer is flexible for various kinds of customers.

“Bolna allows you the freedom to choose any model and has a stack behind it to mold it according to your requirement. It’s a good option for people who want to own some part of the stack, want flexibility in model picking, and want to be able to maintain those products themselves,” Shrivastava told TechCrunch over a call.

Ethos Technologies has priced its initial public offering and the expectation is that it will go public on Thursday, making it one of the first tech IPOs of the year. 

If it lands in its current price range of $18 to $20 a share, it will enter the day valued at $1.26 billion on the high end — raising $102.6 million for itself and about $108 million for its selling shareholders. Should investor interest be high, it could wind up pricing higher. That means a bigger valuation and more money raised.

The company, which offers software to sell life insurance, is backed by Sequoia, Accel, Alphabet’s venture capital arm GV, Softbank, General Catalyst, and Heroic Ventures. Sequoia and Accel are not selling shares in the IPO, the company disclosed.

Ethos was a rising startup star in the pre-AI era, raising one big round after another through 2021. In its early rounds, it was backed by a who’s who of family offices, including those of Will Smith, Robert Downey Jr., Kevin Durant, and Jay Z, the company told TechCrunch in 2018. 

It hit a $2.7 billion valuation in 2021 having raised $400 million by then, most of it that year. It completed only very small fundraises after that, Pitchbook estimates.  

Ethos is profitable and has been for years, its IPO documents disclose. In the nine months ending September 30, Ethos generated almost $278 million in revenue and just under $46.6 million in net income. 

ICE (U.S. Immigration and Customs Enforcement) has now become the No. 3 most-blocked account on Bluesky, after receiving its official verification on Friday, according to third-party trackers. Bluesky users, unsurprisingly, are angry about the government account being hosted on the platform. Many are recommending that others block the account directly or subscribe to a block list that includes all of the U.S. government’s official accounts.

The blocklist was introduced after the White House and other government agencies under the Trump administration signed up for Bluesky last October to post messages blaming Democrats for the government shutdown. The accounts that joined at the time included the Departments of Homeland Security, Commerce, Transportation, the Interior, Health and Human Services, State, and Defense, in addition to the White House itself.

The move made the White House one of the most-blocked accounts on Bluesky, and today it remains in the No. 2 position, just behind Vice President J.D. Vance, per stats shared on the tracking site Clearsky. (The site leverages Bluesky’s API to track which accounts are the most blocked and other blocking activity.)

ICE, however, did not join Bluesky in October. According to Bluecrawler’s Join Date Checker, the account @icegov.bsky.social joined the social network on November 26, 2025.

The account was verified a few days ago according to the independently-run Verified Account Tracker, which suggests that either Bluesky’s team didn’t have enough information to apply the verification checkmark, was somehow unaware of the account’s existence (doubtful!), or was internally debating how to handle the issue. Bluesky hasn’t responded to a request for comment.

One tracker now shows the ICE account as being over 60% of the way to being the most-blocked Bluesky account.

ICE today has many accounts across other social media sites, including X, Instagram, Facebook, YouTube, and LinkedIn. These accounts tend to be verified on platforms that have a verification mechanism, with YouTube being an exception.

The decision from Bluesky to host and verify ICE establishes the social network as one that’s now fitting in more with other, larger social media giants, rather than with the original ethos of the open social web known as fediverse, where the user community is more in control of which accounts gain attention and traction.

The fediverse, which represents a network of independent but interconnected social media platforms, includes apps like Mastodon, Pixelfed, PeerTube, Flipboard, and, to some extent, Instagram Threads, though Meta’s app isn’t fully federated. The U.S. government doesn’t have Mastodon accounts, but users can follow accounts like @potus on Threads from their Mastodon accounts, if they choose.

One reason for avoiding Mastodon, an open source federated app that runs on the ActivityPub protocol, could be its smaller size. But also, any government account joining this network could be easily blocked by individual server operators. This wouldn’t prevent the account from setting up its own server to post to the fediverse, but other communities could refuse to federate (interoperate) with that server, greatly diminishing its reach.

Mastodon’s founder Eugen Rochko, who stepped down as CEO in November, citing burnout, recently posted an anti-ICE message on Mastodon, noting that “Abolish ICE” doesn’t go “nearly far enough” to address the problem in the U.S.

A day later, he announced he was opting his account out of the bridge that connects Mastodon with Bluesky.

Bridging technology, which includes the project known as Bridgy Fed, is meant to allow different decentralized platforms to connect with each other, even if they run different protocols, as is the case with Bluesky, which runs on AT Protocol. Coincidentally, Bridgy Fed today launched a way to add domain blocklists to bridged accounts, which would conceivably allow fediverse users to block the government agencies posting on Bluesky.

Reached for comment, Rochko wouldn’t confirm whether or not ICE’s participation on Bluesky was a factor in his decision to leave the bridge, saying that the decision was a “personal” one.

However, there has often been tension between the fediverse and the atmosphere, or the decentralized social platform that includes Bluesky and other, newer networks and apps like Blacksky, Northsky Social, and more. Because the networks have different approaches to decentralization, they each have their own supporters and critics, some of whom can’t even agree that the networks should be bridged in the first place.