Another Twitter whistleblower is considering taking the stand when Elon Musk faces the company in court over his $44 billion buyout deal later in two weeks — and he may have a few things to say about bots, The Post has learned.
Unlike Peiter “Mudge” Zatko — a Twitter whistleblower who did not mention the word “bots” or “spam” a single time during his two-and-a-half hour Congressional testimony in September — the potential second tipster would focus on an alleged internal study concluding that the site’s bot problem is much larger than Twitter has acknowledged.
The would-be whistleblower, a former Twitter employee, claims to have been involved in an internal report several years ago that supposedly concluded that at least 30% of Twitter’s daily active users are automated spam accounts.
“Twitter execs laughed when they were told about the report and said, ‘We have always had a bot problem,’” the potential whistleblower said in an interview with The Post.
His testimony could be a boon for Musk, who has made Twitter’s alleged bot problem a cornerstone of his legal argument for why he should be allowed to get out of his deal to buy the site.
But Musk still faces a big problem: the potential whistleblower isn’t sure if he wants to talk.
“I haven’t fully decided yet,” the would-be witness said, adding that he’s not sure if he’s ready for the attention that would come with taking the stand at one of the most closely watched trials in years.
The two sides are scheduled to swap witness lists on Wednesday, potentially giving an answer to whether the second whistleblower will appear. While Zatko has been subpoenaed by Musk’s team, the second whistleblower has not received a subpoena.
Representatives for Twitter and Musk declined to comment.
The potential whistleblower said that Twitter’s chosen metric to measure bots — called Monetizable Daily Active Users, or mDAUs — is overly narrow and fails to fully capture the extent of the site’s spam problem. He said Twitter may not be lying when it says that less than 5% of its mDAUs are bots, but takes issue with using mDAU as a measurement in the first place and claims the company has not been transparent with investors.
Twitter says it has 238 million mDAUs — but there are a significantly higher number of overall daily users when one includes the automated fake accounts that have not been disabled, the potential whistleblower said. The spam accounts were identified because they displayed bot-like behavior such as posting every hour on the hour and replying to tweets within five seconds of another user posting them, he said.
“The total number of active accounts was higher than the number we reported publicly,” the whistleblower would-be said.
A source close to Twitter said the company was not aware of the exact study described by the potential whistleblower. The source pointed out that not all automated accounts on the site are spam. So-called “good bots” on the site include accounts like @howsmydrivingny, which automatically looks up traffic violations based on license plate numbers, and @met_drawings, which posts public domain works from the Met’s drawings and prints department on an automated schedule.
Potentially more importantly, the source argued that the potential whistleblower should have no bearing on the five-day trial, which is set to kick off on Oct. 17 in Delaware, since he appears to not allege fraud or inconsistencies in Twitter’s Securities and Exchange Commission filings. Musk’s decision to waive due diligence when he initially agreed to buy Twitter also weakens his argument about bots, legal experts say.
The whistleblower said that he had to delete the report when he left Twitter as part of an agreement to not keep confidential business information.