Proton Releases Major Update to Privacy-Centric AI Chatbot Lumo, Boosting Performance and Accuracy
Proton, renowned for its privacy-focused email service Proton Mail, has introduced a significant update to its AI chatbot, Lumo. Launched in July, Lumo aims to provide users with a secure and private conversational AI experience. Less than a month after its debut, Proton has begun rolling out Lumo 1.1 to both free and subscription users, highlighting notable improvements across multiple performance metrics.
Proton reports that Lumo 1.1 delivers a 170% enhancement in understanding context, enabling the chatbot to interpret and respond to user queries more accurately—especially when dealing with documents and data shared during conversations. Additionally, the update improves code generation capabilities by 40% and enhances multi-step problem-solving by 200%, allowing Lumo to handle complex prompts and utilize appropriate tools more effectively.
One of the initial limitations of Lumo was its less reliable responses about current events. With the latest version, Proton has integrated a better web search functionality, resulting in more precise answers and fewer hallucinations, or false information. Although Proton has not disclosed specific measurement methods, the company shared comparison screenshots illustrating the progress from version 1.0 to 1.1, emphasizing clearer, well-formatted responses.
Maintaining its commitment to privacy, Proton ensures that all interactions with Lumo remain encrypted, with no logs stored or shared. To promote transparency and security, Proton has also open-sourced the code for Lumo’s iOS and Android apps, inviting the developer community to review the security protocols and contribute to ongoing improvements.
For more information on privacy-preserving AI tools, resources such as the official Proton website and AI security best practices are recommended.