Twitter owner Elon Musk’s dictatorial management style risks driving the company headlong into unforced business blunders, content moderation disasters and the degradation of core platform features that help keep vulnerable users safe, according to a former top Twitter official who led the company’s content moderation before abruptly resigning this month.The social media company’s botched rollout of a paid verification feature “is an example of a disaster that slipped through” amid the chaos Musk brought to Twitter, and the prospect of further disasters made it impossible to stay, said Yoel Roth, the company’s former head of site integrity, during an onstage interview with the journalist Kara Swisher Tuesday in his first public appearance since quitting Twitter on Nov. 10.Roth and other colleagues tried to warn Musk of the “obvious” problems in his plan to offer a verified check mark to any user who paid $8 a month. But Musk charged ahead anyway through sheer force of will, leading to a wave of new impostor accounts posing as major brands, athletes and other verified users that soon forced Twitter to suspend the feature.”It went off the rails in exactly the ways that we anticipated,” Roth said.The public reflections of a senior Twitter leader who had close contact with Musk in the raw, early days of his ownership of the company — a period marked by internal tumult and a damaging advertiser revolt — provide the latest evidence of a billionaire CEO who leads by his gut at the expense of virtually everyone else.There was no explosive confrontation with Musk that led to Roth’s resignation, and the episode involving Twitter’s paid verification feature was only one of many factors that drove Roth’s decision to leave, he said. But the experience exemplified the kind of damage Musk’s freewheeling approach can do, Roth added, likening his final weeks at the company to standing before a leaky dam, trying desperately to plug the holes but knowing that eventually something would get past him.In the hour-long interview, Roth warned Musk’s laissez-faire approach to content moderation, and his lack of a transparent process for making and enforcing platform policies, has made Twitter less safe, in part because there aren’t enough staff remaining who understand that malicious actors are constantly trying to game the system in ways that automated algorithms don’t know how to catch.”People are not sitting still,” he said. “They are actively devising new ways to be horrible on the internet.”He urged Twitter users to monitor the functioning of key safety features such as muting, blocking and protected tweets as early warning signs the platform may be breaking down.”If protected tweets stop working, run,” he said.For two weeks after Musk closed his purchase of Twitter, Roth presented himself as a voice of stability and calm at the center of a company undergoing dramatic change. Roth knew that by remaining at the company, Musk was using him to help keep advertisers from abandoning the platform. But Roth also suggested that he and others who did not leave Twitter may have been able to influence Musk and keep him from making damaging unilateral decisions, which he had “multiple opportunities” to do.Even as he spent his initial days in the new regime battling a “surge in hateful conduct on Twitter” apparently meant to test Musk’s tolerance for racism and antisemitism on the platform, Roth sought to reassure the public that Twitter’s trust and safety work continued unhindered.He shared data on the platform’s ongoing enforcement efforts, and downplayed the impact of Twitter’s mass layoffs on its content moderation team, saying the job cuts were less severe in that department compared to the wider organization.As late as Nov. 9, Roth spoke alongside Musk during a public Twitter Spaces event intended to persuade advertisers not to flee the platform. In the hour-long session, which was attended by more than 100,000 listeners, including representatives of Adidas, Chevron and other major brands, Roth waxed optimistic about Twitter’s plans to fight hate speech.The very next day, Roth abruptly resigned, joining a slew of other senior executives including Twitter’s chief privacy officer and chief information security officer.In a subsequent New York Times op-ed, Roth said his reason for leaving came down to Musk’s highly personal and improvisational approach to content moderation. Roth’s essay accused Musk of perpetuating a “lack of legitimacy through his impulsive changes and tweet-length pronouncements about Twitter’s rules.”On Tuesday, Roth said the popular narrative that describes Musk as a villain is wrong and doesn’t reflect his own experiences with him. But, he said, Musk surrounds himself with those who rarely challenge him.Before Musk took over Twitter, Roth wrote down several commitments to himself that would trigger the decision to quit. One limit, he said — one that was never reached — was that Roth would refuse to lie for Musk. Another limit, one that was ultimately reached and drove his decision to resign, was “if Twitter starts being ruled by dictatorial edict rather than by a policy.”Roth’s role at Twitter came under intense scrutiny in 2020 after the company appended a fact-check message to false tweets by then-US President Donald Trump.Tweets that Roth sent in 2016 and 2017 that were critical of President Trump and his supporters were dug up and used to argue that Roth and Twitter were biased against the president.Among Roth’s tweets was one he wrote on Election Day 2016 that read, “I’m just saying, we fly over those states that voted for a racist tangerine for a reason.”Twitter defended Roth at the time, saying, “No one person at Twitter is responsible for our policies or enforcement actions, and it’s unfortunate to see individual employees targeted for company decisions.”When Roth was still working at Twitter in October, Musk was asked about Roth’s old tweets.”We’ve all made some questionable tweets, me more than most, but I want to be clear that I support Yoel. My sense is that he has high integrity, and we are all entitled to our political beliefs,” Musk tweeted.Roth also became the personal face of Twitter, and a target of harassment, after the company decided to suppress a 2020 New York Post story about Hunter Biden, a decision then-CEO Jack Dorsey has since said was a mistake.”It’s widely reported that I personally directed the suppression of the Hunter Biden story. That is not true. It is absolutely, unequivocally untrue,” Roth told Swisher on Tuesday.Roth did not feel removing the content from Twitter was appropriate, he said, but at the time the story seemed to bear the hallmarks of a hack-and-leak information operation.Roth also said Tuesday that, in retrospect, suppressing the Hunter Biden story was a mistake. But he defended Twitter’s other decisions to ban Trump for his activities around the Jan. 6 Capitol riot, as well as a personal account belonging to Georgia Republican Rep. Marjorie Taylor Greene and an account belonging to the satirical website Babylon Bee.All three cases involved obvious violations of Twitter’s publicly accessible, written policies, Roth said, making them a much clearer case for enforcement.Amid the layoffs that have decimated Twitter’s content moderation team, Musk has said he intends to rely much more heavily on crowdsourced fact-checking of tweets to provide context to misleading claims. But Roth said that in doing so, Twitter risks abdicating its responsibility to the public, which should still apply despite it being a private company.Policymakers should require platforms to share data with academics and researchers, he said, preempting privately owned platforms such as Twitter from shirking a duty to transparency.Asked to give a single piece of advice to Musk going forward, Roth paused for the briefest of moments.”Humility goes a really long way,” he said.Twitter did not immediately respond to a request for comment.CNN’s Donie O’Sullivan contributed to this report
Twitter owner Elon Musk’s dictatorial management style risks driving the company headlong into unforced business blunders, content moderation disasters and the degradation of core platform features that help keep vulnerable users safe, according to a former top Twitter official who led the company’s content moderation before abruptly resigning this month.
The social media company’s botched rollout of a paid verification feature “is an example of a disaster that slipped through” amid the chaos Musk brought to Twitter, and the prospect of further disasters made it impossible to stay, said Yoel Roth, the company’s former head of site integrity, during an onstage interview with the journalist Kara Swisher Tuesday in his first public appearance since quitting Twitter on Nov. 10.
Roth and other colleagues tried to warn Musk of the “obvious” problems in his plan to offer a verified check mark to any user who paid $8 a month. But Musk charged ahead anyway through sheer force of will, leading to a wave of new impostor accounts posing as major brands, athletes and other verified users that soon forced Twitter to suspend the feature.
“It went off the rails in exactly the ways that we anticipated,” Roth said.
The public reflections of a senior Twitter leader who had close contact with Musk in the raw, early days of his ownership of the company — a period marked by internal tumult and a damaging advertiser revolt — provide the latest evidence of a billionaire CEO who leads by his gut at the expense of virtually everyone else.
There was no explosive confrontation with Musk that led to Roth’s resignation, and the episode involving Twitter’s paid verification feature was only one of many factors that drove Roth’s decision to leave, he said. But the experience exemplified the kind of damage Musk’s freewheeling approach can do, Roth added, likening his final weeks at the company to standing before a leaky dam, trying desperately to plug the holes but knowing that eventually something would get past him.
In the hour-long interview, Roth warned Musk’s laissez-faire approach to content moderation, and his lack of a transparent process for making and enforcing platform policies, has made Twitter less safe, in part because there aren’t enough staff remaining who understand that malicious actors are constantly trying to game the system in ways that automated algorithms don’t know how to catch.
“People are not sitting still,” he said. “They are actively devising new ways to be horrible on the internet.”
He urged Twitter users to monitor the functioning of key safety features such as muting, blocking and protected tweets as early warning signs the platform may be breaking down.
“If protected tweets stop working, run,” he said.
For two weeks after Musk closed his purchase of Twitter, Roth presented himself as a voice of stability and calm at the center of a company undergoing dramatic change. Roth knew that by remaining at the company, Musk was using him to help keep advertisers from abandoning the platform. But Roth also suggested that he and others who did not leave Twitter may have been able to influence Musk and keep him from making damaging unilateral decisions, which he had “multiple opportunities” to do.
Even as he spent his initial days in the new regime battling a “surge in hateful conduct on Twitter” apparently meant to test Musk’s tolerance for racism and antisemitism on the platform, Roth sought to reassure the public that Twitter’s trust and safety work continued unhindered.
He shared data on the platform’s ongoing enforcement efforts, and downplayed the impact of Twitter’s mass layoffs on its content moderation team, saying the job cuts were less severe in that department compared to the wider organization.
As late as Nov. 9, Roth spoke alongside Musk during a public Twitter Spaces event intended to persuade advertisers not to flee the platform. In the hour-long session, which was attended by more than 100,000 listeners, including representatives of Adidas, Chevron and other major brands, Roth waxed optimistic about Twitter’s plans to fight hate speech.
The very next day, Roth abruptly resigned, joining a slew of other senior executives including Twitter’s chief privacy officer and chief information security officer.
In a subsequent New York Times op-ed, Roth said his reason for leaving came down to Musk’s highly personal and improvisational approach to content moderation. Roth’s essay accused Musk of perpetuating a “lack of legitimacy through his impulsive changes and tweet-length pronouncements about Twitter’s rules.”
On Tuesday, Roth said the popular narrative that describes Musk as a villain is wrong and doesn’t reflect his own experiences with him. But, he said, Musk surrounds himself with those who rarely challenge him.
Before Musk took over Twitter, Roth wrote down several commitments to himself that would trigger the decision to quit. One limit, he said — one that was never reached — was that Roth would refuse to lie for Musk. Another limit, one that was ultimately reached and drove his decision to resign, was “if Twitter starts being ruled by dictatorial edict rather than by a policy.”
Roth’s role at Twitter came under intense scrutiny in 2020 after the company appended a fact-check message to false tweets by then-US President Donald Trump.
Tweets that Roth sent in 2016 and 2017 that were critical of President Trump and his supporters were dug up and used to argue that Roth and Twitter were biased against the president.
Among Roth’s tweets was one he wrote on Election Day 2016 that read, “I’m just saying, we fly over those states that voted for a racist tangerine for a reason.”
Twitter defended Roth at the time, saying, “No one person at Twitter is responsible for our policies or enforcement actions, and it’s unfortunate to see individual employees targeted for company decisions.”
When Roth was still working at Twitter in October, Musk was asked about Roth’s old tweets.
“We’ve all made some questionable tweets, me more than most, but I want to be clear that I support Yoel. My sense is that he has high integrity, and we are all entitled to our political beliefs,” Musk tweeted.
Roth also became the personal face of Twitter, and a target of harassment, after the company decided to suppress a 2020 New York Post story about Hunter Biden, a decision then-CEO Jack Dorsey has since said was a mistake.
“It’s widely reported that I personally directed the suppression of the Hunter Biden story. That is not true. It is absolutely, unequivocally untrue,” Roth told Swisher on Tuesday.
Roth did not feel removing the content from Twitter was appropriate, he said, but at the time the story seemed to bear the hallmarks of a hack-and-leak information operation.
Roth also said Tuesday that, in retrospect, suppressing the Hunter Biden story was a mistake. But he defended Twitter’s other decisions to ban Trump for his activities around the Jan. 6 Capitol riot, as well as a personal account belonging to Georgia Republican Rep. Marjorie Taylor Greene and an account belonging to the satirical website Babylon Bee.
All three cases involved obvious violations of Twitter’s publicly accessible, written policies, Roth said, making them a much clearer case for enforcement.
Amid the layoffs that have decimated Twitter’s content moderation team, Musk has said he intends to rely much more heavily on crowdsourced fact-checking of tweets to provide context to misleading claims. But Roth said that in doing so, Twitter risks abdicating its responsibility to the public, which should still apply despite it being a private company.
Policymakers should require platforms to share data with academics and researchers, he said, preempting privately owned platforms such as Twitter from shirking a duty to transparency.
Asked to give a single piece of advice to Musk going forward, Roth paused for the briefest of moments.
“Humility goes a really long way,” he said.
Twitter did not immediately respond to a request for comment.
CNN’s Donie O’Sullivan contributed to this report
