Elon Musk, who purchased Twitter for a cool $44 billion back in February, is obsessed with bots. He has been for some time now, and in the months since his takeover, he’s been vocal about Twitter’s quest to erase them from the platform, telling his remaining employees that the Musk v Bot battle is the “absolute top priority” for the organization.
And recently, without much evidence, he tweeted several statements to support the notion that Twitter’s made progress — a little over a week ago, he called for bots and trolls to “please attack” him, victoriously claiming that because of Twitter’s recent efforts, the bots were “unable to swim to the top” of his replies.
The next day, he additionally shared a particularly unfortunate meme — captioned with an even more unfortunate tongue-out emoji — insinuating that he was taking Twitter bots to their collective funeral.
While there’s no firm public evidence, though, it could be true that Musk and his team may have made progress against English-language bots. But Twitter isn’t just for English speakers, and a report from The New York Times, supported by an in-depth, weeks-long content analysis by David Thiel, the chief technologist at the Stanford Internet Observatory, appears to conclude that Musk’s Twitter is failing miserably to block content spawned by Chinese-language bots — so much so that bot content “drowned out” Twitter posts related to the rare mass protests that broke out in China last month as the result of the nation’s extreme “zero-COVID” policy.
Information sharing is essential for organizing and driving movements, and social media has played a critical role in modern citizen protests, from Arab Spring to Black Lives Matter to the ongoing protests in Iran. China is notorious for its wide-reaching — and extremely effective — censorship policies, mostly restricting internet users to state-moderated apps like Weibo. Thus, to “protect the content from the reach of government censors, many people turned to Twitter to share what they were seeing,” according to the NYT.
Unfortunately, however, due to an overwhelming amount of bot spam — mostly for porn, escort services, and gambling websites — those efforts were all but mute. Protest content was washed out and impossible to find, ultimately rendering Twitter an ineffective tool of free speech, another one of Musk’s alleged priorities, and community mobilization.
But here’s the catch: the bot issue only presented itself in Chinese-language searches. When reporters at the NYT searched “Beijing” in English, for example, they got tweets from real users. But when they did the same search in Chinese characters, all they got was porn and gambling spam. This process was repeated several times with several other city searches, all to the same effect. (It makes sense to search cities, given that protestors are seeking things like travel locations and shared regional voices.)
Importantly, neither the NYT nor Stanford’s Thiel could find proof that Chinese officials unleashed the bots themselves in an attempt to smother protestors; those kinds of bots have existed on the platform for some time, due to the fact that Chinese censors ban both porn and gambling on their state-moderated apps. There also, as Thiel points out in his analysis, wasn’t really a surge in bot traffic, just a total failure of mitigation.
And to that point, this report says less about China, and a lot more about Twitter itself — how wildly complicated it really is to run, how nuanced the issue of bots really is, and how extremely difficult multinational and multilingual content moderation is to manage. And all of these functions have been made exponentially more difficult, and in some cases likely impossible, in the wake of the company’s utter and complete internal chaos — which has involved major workforce slashes, including of international content moderation teams and the company’s completely disbanded its Trust and Safety Council.
Look, it’s not Musk’s fault that bots exist. They were there before he got there, and he didn’t will porn bot swarms on Chinese protestors.
But it’s clear that Musk’s war on bots certainly isn’t happening everywhere, and too-thin teams, who are reportedly struggling enough as it is, aren’t exactly set up for success. Mass protests rarely happen in China, and Twitter, as an organization, failed to provide the necessary services that its new leadership claims to be hell-bent on providing. Actions have consequences — and here, according to Thiel, the very serious consequences of moderation failure may be long-lasting.
“Given how effective it was at drowning out other people,” Thiel told the NYT. “I wouldn’t be surprised if a government of some kind tries to use this technique in the future.”
READ MORE: How Bots Pushing Adult Content Drowned Out Chinese Protest Tweets [The New York Times]