Moxie Marlinspike 推出注重隱私的 ChatGPT 替代方案

Topics

Latest

AI

Amazon

Apps

Biotech & Health

Climate

Cloud Computing

Commerce

Crypto

Enterprise

EVs

Fintech

Fundraising

Gadgets

Gaming

Google

Government & Policy

Hardware

Instagram

Layoffs

Media & Entertainment

Meta

Microsoft

Privacy

Robotics

Security

Social

Space

Startups

TikTok

Transportation

Venture

More from TechCrunch

Staff

Events

Startup Battlefield

StrictlyVC

Newsletters

Podcasts

Videos

Partner Content

TechCrunch Brand Studio

Crunchboard

Contact Us

Moxie Marlinspike has a privacy-conscious alternative to ChatGPT

If you’re at all concerned about privacy, the rise of AI personal assistants can feel alarming. It’s difficult to use one without sharing personal information, which is retained by the model’s parent company. With OpenAI already testing advertising, it’s easy to imagine the same data collection that fuels Facebook and Google creeping into your chatbot conversations.

A new project, launched in December by Signal co-founder Moxie Marlinspike, is showing what a privacy-conscious AI service might look like. Confer is designed to look and feel like ChatGPT or Claude, but the backend is arranged to avoid data collection, with the open-source rigor that makes Signal so trusted. Your Confer conversations can’t be used to train the model or target ads, for the simple reason that the host will never have access to them.

For Marlinspike, those protections are a response to the intimate nature of the service.

“It’s a form of technology that actively invites confession,” says Marlinspike. “Chat interfaces like ChatGPT know more about people than any other technology before. When you combine that with advertising, it’s like someone paying your therapist to convince you to buy something.”

Ensuring that privacy requires several different systems working in concert.

First, Confer encrypts messages to and from the system using the WebAuthn passkey system. (Unfortunately, that standard works best on mobile devices or Macs running Sequoia, although you can also make it work on Windows or Linux with a password manager.) On the server side, all Confer’s inference processing is done in a Trusted Execution Environment (TEE), with remote attestation systems in place to verify the system hasn’t been compromised. Inside that, there’s an array of open-weight foundation models handling whatever query comes in.

The result is a lot more complicated than a standard inference setup (which is fairly complicated already), but it delivers on Confer’s basic promise to users. As long as those protections are in place, you can have sensitive conversations with the model without any information leaking out.

Confer’s free tier is limited to 20 messages a day and five active chats. Users willing to pay $35 a month will get unlimited access, along with more advanced models and personalization. That’s quite a bit more than ChatGPT’s Plus plan — but privacy doesn’t come cheap.