The Brave browser, focused on the protection of personal data, introduced the Leo AI assistant, which is characterized by an increased level of privacy. After the test period, Leo became generally available to users of the computer version of Brave 1.60 and higher, and will also appear in mobile applications later.

The functionality of Leo is similar to other AI chatbots — it can translate texts, answer questions, summarize information, and generate content. The peculiarity of Leo is that dialogues with him are not recorded, which strengthens the protection of privacy, according to the publication Media Security. Nevertheless, one should be wary of potential inaccuracies in his answers.

Leo uses the Meta Llama 2 model and is available to everyone for free. There’s also a paid Leo Premium subscription for $15 per month that offers an improved Claude Instant assistant based on the Anthropic Claude 2 model.

“AI has a lot of potential, but at the same time, it raises the issue of privacy. We need to find a solution that puts privacy first,” said Brian Bondy, CTO and co-founder of Brave. He noted that the company aims to integrate AI into the online space familiar to users without compromising their privacy.

Brave also promises to provide users of the paid version with a larger selection of AI models, better call quality, priority when the system is under high load, increased speed limits, and access to new products before others. Brian Bondy expressed the opinion that over time, the range of available models will expand, allowing users to choose the most suitable option.