Exploring Privacy-First AI Assistants: Meet Confer
The rise of AI personal assistants has raised significant concerns about privacy. Using these tools often requires sharing personal data, which can be retained by the parent companies behind the models. With OpenAI testing advertising, it’s natural to worry about data collection invading your chatbot interactions, similar to practices seen with Facebook and Google.
A New Player in the AI Sphere: Confer
In December, Moxie Marlinspike, co-founder of Signal, introduced an innovative project called Confer. Designed to function like ChatGPT or Claude, Confer prioritizes user privacy by ensuring data collection is avoided. With its open-source foundation, similar to Signal’s reputation, users can engage in conversations without the concern of their data being used for training or advertising.
A Response to the Nature of Conversations
Marlinspike emphasizes the importance of privacy due to the intimate nature of AI-assisted discussions.
“It’s a form of technology that actively invites confession,” Marlinspike explains. “Chat interfaces like ChatGPT possess more knowledge about individuals than any technology before. Merging that with advertising feels akin to a therapist being paid to convince you to make a purchase.”
Robust Privacy Protections in Action
To uphold user privacy, Confer integrates multiple systems working synergistically.
Firstly, Confer encrypts messages using the WebAuthn passkey system. This standard is particularly effective on mobile devices and Macs running Sequoia, though users can adapt it for Windows or Linux via a password manager. Additionally, all inferencing on the server is conducted within a Trusted Execution Environment (TEE), reinforced with remote attestation systems to verify integrity. Inside this secure setup, a collection of open-weight foundation models manages incoming queries effectively.
The Promise of Secure Conversations
While Confer’s architecture may be more complex than standard setups, it fulfills its core promise to users: confidentiality. As long as these protective measures remain intact, users can confidently engage in sensitive discussions without concerns about information breaches.
Techcrunch Event
San Francisco
|
October 13-15, 2026
Flexible Pricing for Enhanced Features
Confer offers a free tier that limits users to 20 messages a day and five active chats. For $35 a month, users can gain unlimited access, advanced models, and personalized experiences. Though this price tag exceeds that of ChatGPT’s Plus plan, the investment is about securing your privacy.
Here are five FAQs about Moxie Marlinspike’s privacy-conscious alternative to ChatGPT:
FAQ 1: What is Moxie Marlinspike’s alternative to ChatGPT?
Answer: Moxie Marlinspike’s alternative to ChatGPT is a privacy-focused conversational AI designed to prioritize user data protection. Unlike conventional AI models, it aims to minimize data retention and enhance user privacy.
FAQ 2: How does this alternative ensure privacy?
Answer: The AI employs end-to-end encryption, minimized data logging, and runs locally on devices when possible. This means user interactions are not stored or used for training purposes, ensuring sensitive information remains confidential.
FAQ 3: Who can use this AI?
Answer: The AI is available for anyone looking for a privacy-conscious conversational tool. It is beneficial for individuals, businesses, and organizations that prioritize data security and wish to maintain control over their conversations.
FAQ 4: How does it compare to traditional AI models regarding performance?
Answer: While the AI may have certain limitations compared to large-scale models like ChatGPT in terms of training data and breadth of knowledge, it is designed to provide accurate and contextual responses while fully respecting user privacy.
FAQ 5: Is there a cost to use this alternative?
Answer: The pricing model can vary depending on how the application is deployed, either as an open-source project or a commercial solution. While some features may be free, others might come with a subscription fee to ensure ongoing development and support.