Signal co-founder Moxie Marlinspike launched a privacy-focused AI service called Confer in December, offering an alternative to mainstream AI personal assistants like ChatGPT and Claude. Confer aims to provide a similar user experience to these popular platforms while prioritizing data privacy and avoiding the data collection practices common among large tech companies.

The impetus behind Confer stems from growing privacy concerns surrounding AI chatbots. These services often require users to share personal information, which is then retained by the parent company and potentially used for purposes like targeted advertising. Marlinspike argues that the intimate nature of AI interactions necessitates stronger privacy safeguards. "Its a form of technology that actively invites confession," Marlinspike said. "Chat interfaces like ChatGPT know more about people than any other technology before. When you combine that with advertising, its like someone paying your therapist to convince you to buy something."

Confer's architecture is designed to prevent data collection. Conversations are not used to train the AI model or target advertisements, as the host never gains access to the data. This is achieved through an open-source backend, similar to Signal's approach, which allows for public scrutiny and verification of its privacy protections.

The rise of AI personal assistants has sparked debate about the balance between convenience and privacy. While these tools offer numerous benefits, such as personalized assistance and information retrieval, they also raise concerns about data security and potential misuse of personal information. OpenAI, the company behind ChatGPT, is already testing advertising, raising the possibility of data collection practices similar to those employed by Facebook and Google.

Confer represents an attempt to address these concerns by offering a privacy-conscious alternative. By prioritizing data protection and transparency, Marlinspike hopes to demonstrate that it is possible to develop AI services that respect user privacy without sacrificing functionality. The project is still in its early stages, and its long-term success will depend on its ability to attract users and compete with established players in the AI market. However, it signals a growing demand for privacy-focused AI solutions and could influence the development of future AI technologies.