Bluesky users report AI bots, disinformation, and copycat accounts

Bluesky has seen its user base soar since the U.S. presidential election, boosted by people seeking refuge from Elon Musk’s X, which they view as increasingly leaning too far to the right given its owner’s support of President-elect Donald Trump, or wanting an alternative to Meta’s Threads and its algorithms.

The platform grew out of the company then known as Twitter, championed by its former CEO Jack Dorsey. Its decentralized approach to social networking was eventually intended to replace Twitter’s core mechanic. That’s unlikely now that the two companies have parted ways. But Bluesky’s growth trajectory — with a user base that has more than doubled since October — could make it a serious competitor to other social platforms.

But with growth comes growing pains. It’s not just human users who’ve been flocking to Bluesky but also bots, including those designed to create partisan division or direct users to junk websites.

The skyrocketing user base — now surpassing 25 million — is the biggest test yet for a relatively young platform that has branded itself as a social media alternative free of the problems plaguing its competitors. According to research firm Similarweb, Bluesky added 7.6 million monthly active app users on iOS and Android in November, an increase of 295.4% since October. It also saw 56.2 million desktop and mobile web visits, in the same period, up 189% from October.

Besides the U.S. elections, Bluesky also got a boost when X was briefly banned in Brazil.

“They got this spike in attention, they’ve crossed the threshold where it is now worth it for people to flood the platform with spam,” said Laura Edelson, an assistant professor of computer science at Northeastern University and a member of Issue One’s Council for Responsible Social Media. “But they don’t have the cash flow, they don’t have the established team that a larger platform would, so they have to do it all very, very quickly.”

To manage growth for its tiny staff, Bluesky started as an invitation-only space until it opened to the public in February. That period gave the site time to build out moderation tools and other distinctive features to attract new users, such as “starter packs” that provide lists of topically curated feeds. Meta recently announced that it is testing a similar feature.

Compared to the bigger players like Meta’s platforms or X, Bluesky has a “quite different” value system, said Claire Wardle, a professor at Cornell University and an expert in misinformation. This includes giving users more control over their experience.

“The first generation of social media platforms connected the world, but ended up consolidating power in the hands of a few corporations and their leaders,” Bluesky said on its blog in March. “Our online experience doesn’t have to depend on billionaires unilaterally making decisions over what we see. On an open social network like Bluesky, you can shape your experience for yourself.”

Because of this mindset, Bluesky has achieved a scrappy underdog status that has attracted users who’ve grown tired of the big players.

“People had this idea that it was going to be a different type of social network,” Wardle said. “But the truth is, when you get lots of people in a place and there are eyeballs, it means that it’s in other people’s interests to use bots to create, you know, information that aligns with their perspective.”

Little data has emerged to help quantify the rise in impersonator accounts, artificial intelligence-fueled networks and other potentially harmful content on Bluesky. But in recent weeks, users have begun reporting large numbers of apparent AI bots following them, posting plagiarized articles or making seemingly automated divisive comments in replies.

Lion Cassens, a Bluesky user and doctoral candidate in the Netherlands, found one such network by accident — a group of German-language accounts with similar bios and AI-generated profile pictures posting in replies to three German newspapers.

“I noticed some weird replies under a news post by the German newspaper ‘Die Ziet,'” he said in an email to The Associated Press. “I have a lot of trust in the moderation mechanism on Bluesky, especially compared to Twitter since the layoffs and due to Musk’s more radical stance on freedom of speech. But AI bots are a big challenge, as they will only improve. I hope social media can keep up with that.”

Cassens said the bots’ messages have been relatively innocuous so far, but he was concerned about how they could be repurposed in the future to mislead.

There are also signs that foreign disinformation narratives have made their way to Bluesky. The disinformation research group Alethea pointed to one low-traction post sharing a false claim about ABC News that had circulated on Russian Telegram channels.

Copycat accounts are another challenge. In late November, Alexios Mantzarlis, director of the Security, Trust and Safety Initiative at Cornell Tech, found that of the top 100 most followed named individuals on Bluesky, 44% had at least one duplicate account posing as them. Two weeks later, Mantzarlis said Bluesky had removed around two-thirds of the duplicate accounts he’d initially detected — a sign the site was aware of the issue and attempting to address it.

Bluesky posted earlier this month that it had quadrupled its moderation team to keep up with its growing user base. The company also announced it had introduced a new system to detect impersonation and was working to improve its Community Guidelines to provide more detail on what’s allowed. Because of the way the site is built, users also have the option to subscribe to third-party “Labelers” that outsource content moderation by tagging accounts with warnings and context.

The company didn’t respond to multiple requests for comment for this story.

Even as its challenges aren’t yet at the scale other platforms face, Bluesky is at a “crossroads,” said Edward Perez, a board member at the nonpartisan nonprofit OSET Institute, who previously led Twitter’s civic integrity team.

“Whether BlueSky likes it or not, it is being pulled into the real world,” Perez said, noting that it needs to quickly prioritize threats and work to mitigate them if it hopes to continue to grow.

That said, disinformation and bots won’t be Bluesky’s only challenges in the months and years to come. As a text-based social network, its entire premise is falling out of favor with younger generations. A recent Pew Research Center poll found that only 17% of American teenagers used X, for instance, down from 23% in 2022. For teens and young adults, TikTok, Instagram and other visual-focused platforms are the places to be.

Political polarization is also going against Bluesky ever reaching the size of TikTok, Instagram or even X.

“Bluesky is not trying to be all things to all people,” Wardle said, adding that, likely, the days of a Facebook or Instagram emerging where they’re “trying to keep everybody happy” are over. Social platforms are increasingly splintered along political lines and when they aren’t — see Meta’s platforms — the companies behind them are actively working to de-emphasize political content and news.

The Associated Press receives support from several private foundations to enhance its explanatory coverage of elections and democracy. See more about AP’s democracy initiative here. The AP is solely responsible for all content.

—Ali Swenson and Barara Ortutay, Associated Press

https://www.fastcompany.com/91252248/bluesky-ai-bots-disinformation-copycat-accounts?partner=rss&utm_source=rss&utm_medium=feed&utm_campaign=rss+fastcompany&utm_content=rss