Elon Musk taps the Twitter whistleblower for help getting out of the deal

With the Twitter trial date rapidly approaching, Elon Musk's legal team sent a subpoena to former Twitter head of security Peiter "Mudge" Zatko, who filed a whistleblower complaint against the company that was made public last week. In the complaint, Zatko alleges that he witnessed "egregious deficiencies, negligence, willful ignorance, and threats to national security and democracy" within Twitter, which he says tried to hide its messy inner workings from regulators and investors.

Zatko, a well-respected security researcher, joined the company in 2020 after hackers gained access to a cluster of high-profile Twitter accounts — Joe Biden and Elon Musk among them — to promote a cryptocurrency scam. He was fired in January by Parag Agrawal, who replaced Jack Dorsey as the company's chief executive.

Musk's team is seeking a deposition and a broad swath of documents from Zatko, hoping to bolster its case before the October 17 trial in Delaware’s Chancery Court. Zatko also received a subpoena from Congress in light of the whistleblower complaint and will appear before the Senate Judiciary Committee next month.

In the filing, Musk's legal team asks for all sorts of things, including any documents or communications related to the impact of spam on Twitter's business and its use of mDAU (more on that shortly) as a "key metric." But they're casting a wide net, and have also requested anything about security vulnerabilities, foreign spies working at Twitter, or Twitter's "attempts to hide its security vulnerabilities from investors, regulators, and/or the public."

Beyond his security concerns, Zatko veered outside of his area of expertise to back Musk's concerns about the number of bots on Twitter. As Musk tries to kill his agreement to buy Twitter for $44 billion, the world's richest man has repeatedly pointed to the platform's problem with bots, claiming that the company misrepresents the total amount of spam and nonhuman accounts on the platform to portray itself in a more flattering light.

Musk is clearly scrambling for a reason out of the deal at this point — after all, he vowed to "defeat the spam bots or die trying" back in April — so the whistleblower complaint provides some fresh fodder that his legal team can try to leverage as it makes the case he should be able to walk away. But just because Musk wants to enlist Zatko to back up his claim that Twitter somehow misled him doesn't mean the bot bits in the whistleblower complaint will actually have any bearing on the situation.

Part of the confusion is that Musk has accused Twitter of falsely claiming that the total percentage of bots on the platform is less than 5%. In reality, Twitter only uses the 5% figure when talking about the percentage of bots within a specific chunk of users: something called mDAU, which stands for "monetizable daily active users." The company says that less than 5% of the mDAU is made up of bots.

Twitter says that it actively filters bots and spam accounts out of its mDAU metric, which it created to give advertisers a sense of how many human beings could be reached with ads. It's all pretty confusing, mostly because the metric is something kind of weird and nonstandard that Twitter came up with, and it's made more confusing by Musk's frequent conflation of the two metrics (the user base at large versus the user base with most spam accounts filtered out, at least in theory).

In his letter notifying the company that he wanted to terminate the deal, Musk commented on the mDAU metric specifically, claiming that his suspiciously timed suspicions amount to an adverse event that should allow him to bail on the deal. To make matters even more confusing, Twitter has previously admitted to miscalculating mDAU.

Relying on mDAU instead of a more standard metric is unusual, but that isn't really what's at issue here. Arguably, none of this bot stuff is at issue at all — it really depends on what a judge decides should fly in Musk's quest to shirk his binding commitment to buy Twitter. And while Zatko's report casts doubt on the use of mDAU as a metric and a bunch of more substantial stuff on the security side, it also backs up Twitter's claim that the company keeps spam out of the mDAU because the whole point of the mDAU is to give advertisers an idea of how many humans might interact with ads. Twitter arguably doesn't really have any reason to inflate this number by juicing it with bots because that would make it look like ads perform worse on the platform (because bots aren't interacting with ads).

The Twitter whistleblower isn't a bot expert, and again, the bot stuff is a Hail Mary from the Musk camp, but Zatko's involvement could support Musk in other ways. There's a world in which Musk's legal team could leverage Zatko's more serious concerns — like that foreign governments were easily able to infiltrate the company or that Twitter misled regulators about its security practices — to argue that Musk should be allowed out of the deal. Based on the wide-ranging requests that Musk's legal team is making, they seem to be quickly pressing forward with a see-what-sticks approach.

Unfortunately for Musk, that approach itself might not stick. Last week, a judge rejected his request for another massive swath of data from Twitter as "absurdly broad" and suggested that the company's cooperation to date was sufficient.