Musk stiffing Google could unleash yet more abuse on Twitter, report says

In what might be another blow to the stability of Twitter's trust and safety efforts, the company has allegedly stopped paying for Google Cloud and Amazon Web Services (AWS), which host tools that support the platform's safety measures, Platformer reported this weekend.

According to Platformer, Twitter relies on Google Cloud to host services "related to fighting spam, removing child sexual abuse material, and protecting accounts, among other things." That contract is up for renewal at the end of this month after being negotiated and signed prior to Elon Musk's takeover. Since "at least" March, Twitter has been pushing to renegotiate the contract ahead of renewal—unsurprisingly seeking to lower costs, Platformer reported.

But now it's unclear if the companies will find agreeable new terms on time or if Musk already intends to cancel the contract. Platformer reported that Twitter is rushing to transition services off the Google Cloud Platform and seemingly plans to drop the contract amid failed negotiations.

Should the companies remain at odds, Twitter could end up canceling licensing to display tweets in Google search results.

Twitter has also stopped paying Amazon, Platformer reported, leading to threats from Amazon to stop paying for Twitter advertising.

AWS told Ars it has no comment on Platformer's report. Twitter and Google did not immediately respond to Ars' request for comment.

Twitter’s escalating trust and safety concerns

The headlines about Twitter's trust and safety department this year have not exactly been reassuring.

Not only did both its brand safety chief and its trust and safety chief recently resign, but Twitter has also seen hate speech surge. Spam bots remain a problem, and propaganda accounts are thriving. Perhaps most concerningly, Twitter failed to remove child sexual abuse materials (CSAM) while hosting buyers and sellers, researchers revealed this month.

Twitter has also seemingly become incentivized to turn a blind eye to malicious activity from its paid subscribers—failing to remove 99 percent of hate speech posted by Blue Verified users. A barrage of hate speech during the West Hollywood Pride parade got so overwhelming—partly because Twitter allegedly did not respond to reports of hate speech—that the Los Angeles County district attorney's office shut down its Twitter account last week, the Los Angeles Times reported.

Now, Platformer has reported that a Twitter service called Smyte—an automated anti-abuse and anti-harassment tool that was previously operating on Google Cloud Platform (GCP)—will potentially shut down on June 30. This could lead to a flood of spam bots and CSAM on Twitter as bots and content could fail to be removed.

The last time Smyte became unstable was last December. That was right around the same time when The New York Times reported a rise in hate speech on Twitter, just before the company abruptly disbanded its Trust and Safety Council.

It seems the only constant regarding trust and safety since Musk took over Twitter has been instability.

According to Platformer, Twitter's ability to automatically detect and investigate bad actors, gratuitous gore, violent extremism, and other abuses all previously depended on the GCP to run.

No end to cost-cutting in sight

Companies like Amazon and Google are unlikely to discuss details of client relationships publicly, so it's hard to confirm whether Twitter actually stopped paying its AWS or GCP bills or how that decision could impact services.

But there is evidence that failing to pay bills has been part of Musk's cost-cutting strategy at Twitter, suggesting that could be the case for major vendors like Amazon or Google. In December, Twitter was sued after the company allegedly stiffed a software vendor and refused to pay $8 million left on a contract. At least six companies had sued Twitter, alleging non-payment as of February this year, CNBC reported. Another lawsuit hit Twitter in January when Musk apparently stopped paying rent at Twitter headquarters.

Former Twitter employees told Platformer that Twitter was wasting money investing in Google Cloud services that it wasn't using, so it's possible that Twitter could be delaying payment until Musk has assessed which services can be cut. Its trust and safety team spent more than $10 million annually on Google Cloud, Platformer reported.

Musk is no longer CEO of Twitter. That job is now Linda Yaccarino's, while Musk remains an executive chair and chief technology officer. For Musk, however, the focus remains on developing Twitter into a sleeker, more profitable product while operating on a budget. With Twitter ad sales reportedly still plunging, the need for cost-cutting seemingly remains, and trust and safety has historically been an area where Musk has made steep cuts.

The @TwitterSafety account hasn't tweeted any updates since April. The only recent move that Twitter seems to have taken to enhance Twitter trust and safety was to cut off free users from DMing people who don't follow them. Announced earlier this month by @TitterDaily, Twitter has decided that only Verified Blue accounts can access this privilege, promising that "this would nearly eliminate DM bot spam at a stroke." Previously, any user could set their DMs as open to all.

Musk responded to that thread, confirming that this update would arrive this week and warning about another way that Twitter could become less trustworthy as a platform.

"As I’ve said many times, it is increasingly difficult to distinguish between AI bots," Musk tweeted. "Soon, it will be impossible."

Musk said one way to tamp down on AI bots spamming users is to charge Twitter users to DM strangers.

"The only 'social networks' that survive will be those that require verification," Musk tweeted. "The payment system is a means of verification that increases bot cost by ~10,000X."