Government censorship, scam accounts, celebrity impersonation, propaganda and a user revolt over a journalist accused of demonizing trans people — these are just some of the hurdles faced by the nascent social media platform Bluesky. But they’re also signs of its rapid growth since the election.
As Bluesky attracts more activity, particularly from those fleeing Elon Musk’s X, it is facing the price of success: tough moderation decisions and a growing number of bad actors.
Nearly 25 million people have signed up for the platform, according to metrics site ClearSky. While that number is still a fraction of X’s user base, Bluesky’s controversies and challenges are gaining mainstream attention — a sign of its increasing cultural relevance.
Last week, the platform faced its most significant controversy yet: user backlash against journalist and media personality Jesse Singal joining the platform. Singal, who has reported on people who reversed their gender transitions among other topics pertaining to trans people, hosts a podcast that critiques perceived left-wing biases in media. LGBTQ nonprofit GLAAD included Singal in its Accountability Project documenting “anti-LGBTQ rhetoric and discriminatory actions” by public figures.
Singal’s presence on Bluesky, considered a hub for trans social media users, sparked a petition signed by over 25,000 people calling for his removal under the platform’s moderation policies. Singal is now the most blocked person on Bluesky. Bluesky didn’t address Singal explicitly in its response to the petition, but posted that it does not take down accounts based on activity off the platform. Bluesky didn’t respond to an NBC News question about the controversy around Singal, who has reported receiving death threats he says the platform has not adequately addressed.
Despite the uproar spilling onto other social media platforms like X and attracting attention from mainstream figures like Lizzo, most Bluesky users appear to be sticking with the platform.
Beyond cultural flash points, Bluesky is also grappling with a range of brass-tacks moderation issues that come as the price of popularity.
In one 24-hour period this November, the company announced it had received a record 42,000 moderation reports. “We appreciate your patience as we dial our moderation team up to max capacity and bring on new team members to support this load,” Bluesky said.
After Brazil’s top court suspended access to X in August, Bluesky saw a surge in Brazilian users, including “stan” communities devoted to particular artists. That led to an influx of copyright complaints, leaving Bluesky’s slim moderation team to sift through which posts to keep and remove, said Aaron Rodericks, Bluesky’s head of trust and safety, in an interview.
“Brazilians love doing memes. They’re a fantastic user base. But our copyright requests went through the roof,” said Rodericks, who previously helped lead Twitter’s trust and safety team.
Scammers have also followed the migration of social media users from X to Bluesky, Sean Gallagher, a principal threat researcher at the cybersecurity company Sophos, told NBC News. “Over the past few weeks, there has been a rapid increase in scam activity,” he said.
Many are romance and “pig butchering” scammers who follow the same playbook they use on other social media sites: posing as romantic prospects in the hopes of establishing a fake relationship with a victim they can later exploit for money, Gallagher said.
So far, Bluesky “has been aggressive in shutting down fraudulent accounts,” Gallagher said, and appears to respond well to people reporting suspicious accounts.
Rodericks said even as Bluesky has experienced growing pains, they’re still not at the large scale seen on other major social media platforms. “There’s things that come with hundreds of millions of users, in terms of harms, that we’re not seeing yet.”
But that hasn’t stopped some critics from zeroing in on the issues.
Far-right X account Libs of TikTok, which is known for targeting trans people online, criticized Bluesky on Thursday, posting screenshots of articles written about the platform — one written by Singal about the death threats he said he received, and another about child sexual abuse material (CSAM) on the platform.
Bluesky has seen only a handful of instances of people posting CSAM, Rodericks said, when compared to the levels seen on other platforms, but the numbers are increasing.
In 2023, while the site could only be accessed through invitation and before it had opened to the public, there was only one verified complaint to the National Center for Missing & Exploited Children, the U.S. organization tasked with tracking online exploitation. This year, Rodericks said, there have been over 830 CSAM cases, each of which the company investigates manually. Most major social media platforms report much higher rates of CSAM. In 2023, X reported 273,416 instances of CSAM on its platforms, according to the center. Instagram reported 11,430,007.
Posting CSAM prompts an immediate and full ban on the site, Rodericks said.
The site also has its share of fake celebrity accounts, a problem that continues to plague larger platforms too. Last month, Bluesky began labeling fake celebrity accounts with “impersonation” if they didn’t label themselves as satire or fan accounts, ending the reign of a handful of accounts that registered famous people’s usernames soon after the platform launched.
While it doesn’t appear that there is a vast ecosystem of celebrity impersonators aimed at scamming users, NBC News easily found a handful of unlabeled but clearly fake accounts for celebrities like Ellen DeGeneres, Oprah Winfrey and Cristiano Ronaldo. Representatives for those three didn’t respond to requests for comment.
Bluesky’s approach to verifying that accounts are who they say they are — a problem that has plagued X in particular since Musk’s changes last year — is both unique and more technically difficult than other sites’. It treats every account as a website and lets anyone who owns a web domain register a Bluesky account with that URL. So instead of Bluesky verifying people’s identity individually, it lets anyone with an official website verify themselves and other users.
“We are getting feedback that some people are finding it too technically challenging, because people are just used to usernames. So the question is whether through enough user education, etc., we can meet those needs through domain verification, or if we have to add other aspects in addition to that,” Rodericks said.
Given Bluesky’s limited resources, Rodericks said, it doesn’t have immediate plans to label state-controlled news agencies from authoritarian governments, as Meta does and X did before Musk ended that practice. There are at least three Bluesky accounts that use the names of and only post stories from state-controlled news agencies — China Daily and People’s Daily from China, and RT from Russia — but none has more than a few hundred followers. None of the three news agencies responded to emails asking if those accounts were authentic.
Disinformation researchers routinely identify account swarms, particularly pushing messages sympathetic to the Russian, Chinese and Iranian governments, on major social media platforms, although they rarely get significant traction. Often, those investigations start when the FBI alerts a social media platform to U.S. intelligence that it’s being exploited by an adversary to the U.S.
To date, the FBI has not made that kind of outreach to Bluesky, Rodericks said. It also doesn’t have a dedicated internal team to hunt that kind of propaganda, but instead relies on user tips, automated sensors for inauthentic account behavior, and a volunteer moderator army managed through a nonprofit called Independent Federated Trust & Safety to identify and take down activity that violates its terms of service.
While there have not been clearly documented mass propaganda campaigns on Bluesky so far, researchers have seen some signs.
“All social media platforms are used by adversaries,” Lisa Kaplan, the CEO of Alethea, a company that tracks coordinated online messaging and propaganda campaigns, told NBC News. “We’ve seen evidence of what is likely state actor activity on the platform. That said, it’s early and we’ve not yet seen a successful, coordinated effort.”
“In general, we’re still very small fries, so people aren’t dedicating resources towards us from a nation-state kind of front,” Rodericks said.
Bluesky has been blocked by at least two censorship-prone governments. According to the Great Firewall, a program that tracks internet censorship in China, the country began blocking Bluesky ahead of June 4, the anniversary of the Tiananmen Square protests, a common spur for censorship in the country. Separately, Pakistan briefly blocked Bluesky in November, according to NetBlocks, a company that tracks global internet blocking. Spokespeople for both countries’ consulates didn’t respond to requests for comment.
Every social media platform has to walk the line between obeying the laws of every country it wants to operate in and deciding to resist potentially unfair government requests to block posts or turn over user information. But China and Pakistan didn’t prompt a debate, Rodericks said.
“Neither government has done any type of outreach or communication to Bluesky,” he said. “From a principles perspective, we of course believe in a free and open internet, so we’re going to do our best to support the ability of citizens to access the community’s information. But as a small player in this space, we’ll do what we can to advance our objective balance against countries being able to control their own internet.”