Signal founder says Confer’s private AI tech is coming to Meta AI

Signal founder Moxie Marlinspike just came out and said that Confer’s private AI technology is heading straight into Meta AI. It seems like a pretty surprising team-up when you think about it. Here you have one of the most respected voices in privacy joining forces with a company that people rarely trust on that front.

In a blog post, Marlinspike explained that Confer’s mix of private AI and end-to-end encryption is coming to Meta’s AI tools. In fact, when someone replied on X asking about the iOS app release, he simply answered that they are just waiting on Apple review.

Confer is not just some random startup idea. Back in January, TechCrunch called it a real privacy-focused option compared to tools like ChatGPT or Claude. The whole point is that your chats stay private because the company running it never gets to see them. So nothing gets used for training models or ads. It runs the actual processing inside something called a Trusted Execution Environment and even uses remote attestation to prove the setup has not been tampered with. 

Meta’s AI features on WhatsApp have already drawn privacy scrutiny. The Verge reported last year that AI requests on WhatsApp were being handled on Meta’s servers by default, while Wired noted that users were uneasy because chats with Meta AI did not have the same protections as standard end-to-end encrypted messages, prompting Meta to introduce its “Private Processing” system as a privacy fix.

So yes, Marlinspike’s statement looks like genuine news. But it is still early days, and there is plenty we still do not know. We have not heard much yet about exactly where this will appear first, how much of the system actually stays protected, whether Meta can still see any prompts in some way, or if any part of it will get outside audits. Those details matter a lot. They basically decide whether the whole thing works.

The interesting thing here is that this news comes just days after we reported about Instagram dropping support for end-to-end encrypted direct messages starting May 8. The platform claims E2EE was not being used by many, and thus, they decided to stop supporting it altogether. 

So on one hand, Meta is stepping back from encryption in a product people already have. On the other it is suddenly tying itself to a privacy-first AI effort led by the guy behind Signal.

There is also a longer history here. The New York Times ran a piece in February pointing out Meta’s repeated privacy problems over the years, including huge settlements tied to facial recognition. They also noted that the company loosened some of its internal privacy checks earlier this year.

We also recently covered how Meta AI smart glasses have been sending intimate videos to human moderators. So it’s clear Meta has to change how it moderates and processes such information and media.

At the end of the day, the story boils down to this. Marlinspike says Confer’s private AI tech is coming to Meta AI. The bigger issue is whether Meta can actually make users believe it after everything else they have heard before.

Featured image generated with AI

We stand out from the tech-media crowd because we break news stories; we mainly bring you stuff that you won’t find anywhere in the mainstream tech media. Our stories have been picked up by some of the world’s most popular websites and media outlets—more info is available here.