Roblox stock fell a whopping nine percent after the publication of a report that claims the software is a “pedophile hellscape,” alongside figures that it claims suggest the company is massively exaggerating the number of people using it.
For those paying close attention, games-platform-cum-creation-tool Roblox has seemed like a place you might wanna keep your kids away from for years. However, for most parents, it’s routinely assumed to be a safe online playground. The colossally successful portal for wildly popular user-created games has repeatedly been accused of doing far too little to protect its young users from exploitation, and from sexual predators. This is now underlined by a report from short-selling investigative firm Hindenburg Research.
As reported by Fortune, Hindenburg Research has uploaded its latest investigation under the title, “Roblox: Inflated Key Metrics For Wall Street And A Pedophile Hellscape For Kids.”
The accusations within are, as the title might suggest, incredibly incendiary. This begins with suggestions that Roblox Corp’s share prices are inflated via “growth metrics it presents to Wall Street” that don’t represent reality. The company, which has made a loss every quarter since going public in 2021, is accused of “lying to investors, regulators and advertisers about the number of ‘people’ on its platform, inflating the key metric by 25-42%+.” It then adds, “We also show how engagement hours, another key metric, is inflated by an estimated 100%+.”
These purported figures are based on comparing the numbers Roblox states publicly to those provided by former employees who claim internal numbers are markedly different, accounting for multiple accounts and bots. The report goes as far as to describe this as being like “two sets of books for counting users: one for internal business decisions…and one used by the financial team that reports higher metrics to investors.”
Accusations of grossly inflated player numbers go a lot further. Hindenburg says its research has shown that Roblox’s second-most popular game, Blox Fruits, is “dominated by traffic from Vietnam,” where extremely popular Facebook groups were advertising methods to run over 20 Roblox tabs at a time, something players might be incentivized to do because they can sell the fruits of their labor for in-game currency.
Hindenburg says it hired a technical consultant who “monitored the top ~7,200 Roblox games across ~2.1 million Roblox servers, collecting 297.7 million rows of real-time player data.” From this extensive data, the investigative short seller concludes that rather than the average user playing 2.4 hours a day as Roblox’s data indicates, it was in fact closer to around 22 minutes. The report suggests this is in part due to bots, and in part thanks to its claim that Roblox “incentivizes developers to create ‘AFK’ games that artificially inflate engagement.”
Following on from investigations by Wired and the excellent People Make Games in 2021, Hindenburg saves its most provocative language for the topic of child protection. It uses the words, “our in-game research revealed an X-rated pedophile hellscape, exposing children to grooming, pornography, violent content and extremely abusive speech.”
The report makes specific reference to a 2024 report by the National Center on Sexual Exploitation titled, “Roblox Treats Child Protection Like A Game.”
Here the report’s evidence becomes more anecdotal than analytical, but it still finds some damning evidence that suggests a broad lack of care when it comes to protecting children. Multiple accounts registered to variations on the name “Jeffrey Epstein” had usernames like “@igruum_minors” and “@RavpeTinyK1dsJE.” The most shocking reported element reads,
After we found a username, we listed our age as “under 13” to see if children are being exposed to adult content. By merely plugging ‘adult’ into the Roblox search bar, we found a group called “Adult Studios” with 3,334 members openly trading child pornography and soliciting sexual acts from minors.
We tracked some of the members of “Adult Studios” and easily found 38 Roblox groups—one with 103,000 members—openly soliciting sexual favors and trading child pornography.
The “Adult Studios” group has been locked since the report’s publication.
The report goes on to detail games available in Roblox that seem like they should easily be spotted and blocked, with names like Escape to Epstein Island and Run From Diddy Simulator. Hindenburg also found games with a million visits like Beat Up Homeless Outside 7/11 Simulator and Beat Up The Pregnant, the latter described as one in which “users hacked pregnant women to death in a Wal-Mart parking lot with machetes or killed them with frying pans or a selection of guns.”
Further,
Posing as a child in Roblox’s “therapy” experience, our therapist introduced himself as a “rapper with only one p”. We were advised to run away from home and that he would come pick us up so we could move into his basement in exchange for paying rent with our body.
The recurring theme in the report is that when registering accounts as children aged 8 or 9, there were no restrictions to adult-themed groups, or even groups that had been repeatedly moderated for “trading underage pornography.”
Roblox Corporation has responded to Hindenburg’s report overnight. The company titles its reply, “Roblox Refutes Misleading Claims In Hindenburg Report” (pdf), and begins with wildly irrelevant statements like, “Every day, tens of millions of users of all ages have safe and positive experiences on Roblox, abiding by the company’s Community Standards.”
This is unquestionably true, but it doesn’t refute any claims made by Hindenburg. It’s a bit like a car manufacturer responding to reports that one of its cars keeps exploding by saying, “Millions of customers don’t blow up every day.” It then reiterates the company’s commitment to “proactive and preventative safety measures,” which one might argue aren’t something to boast about in the face of thousands of words of detailed evidence about how they aren’t working.
The response then moves on to asserting that Hindenburg’s financial claims are “misleading.” It points out that as short sellers, Hindenburg has a financial incentive here, and goes on to talk about “cash bookings” and “cash flow,” which it claims the report leaves out because the “facts simply don’t support their agenda.”
There’s then a section refuting the claims that Roblox reports deceptive user figures to investors, by quoting a section from their own accounts that openly explains how the user numbers are calculated, explicitly stating, “our DUAs [Daily Active Users] are not a measure of unique individuals accessing Roblox.” One might suggest that “Daily Active Users” isn’t a brilliant term to use in this light, but more significantly, it fails to recognize that the specific claim made by Hindenburg’s report was that the company uses different internal figures for its own business decisions than those it shares publicly. There is no mention of this in the reply at all.
It’s a very peculiar response, given the severity of the claims made by Hindenburg Research, especially with respect to the severe failings encountered when it comes to child protection. Simply stating that Roblox “has invested heavily in its Trust & Safety efforts,” with no mention that there may still be room for improvement, or any expression of regret over the serious failings Hindenburg says it’s discovered, to me suggests ambivalence at best.
Roblox Corporation is the company whose studio head, Stefano Corazza, said earlier this year that the abysmal fees paid to young users who create the games in Roblox are “a gift” for “15-year-olds, in Indonesia, living in a slum.”
.