Moxie Marlinspike, the cryptographer who created Signal and co-designed the encryption protocol that now protects billions of WhatsApp messages, announced this week that his encrypted AI platform Confer will bring its privacy technology to Meta AI.

The Problem He's Trying to Solve

AI chatbots have become some of the most data-rich systems ever built. People share their medical concerns, financial details, and unfiltered thoughts with them daily — none of which is protected by encryption. As Marlinspike put it in his blog post: "It is shared with AI companies, their employees, hackers, subpoenas, and governments. As is always the case with unencrypted data, it will inevitably end up in the wrong hands."

Confer + Meta AI

Marlinspike launched Confer in early 2026 as a privacy-first AI assistant — built so that no one, not even Confer itself, can access user conversations. The platform runs on open-weight models and uses cryptographic techniques to keep inference private.

His announcement: Confer's privacy architecture will now underpin Meta AI. Confer continues to operate independently, but its encryption layer will serve as the privacy foundation for Meta's AI products going forward.

History Repeating

This isn't Marlinspike's first collaboration with Meta on encryption. In 2016, he worked with WhatsApp — also owned by Meta — to roll out end-to-end encryption to over a billion accounts simultaneously. That move helped make E2EE mainstream. His goal now is to do the same for AI chat.

If the integration succeeds, it would represent the most significant privacy upgrade to a major AI platform since ChatGPT launched.