Proton, creator of secure productivity tools, announced significant upgrades to its privacy-centric AI assistant Lumo on Tuesday. The new 2.0 iteration introduces transformative features while maintaining the company’s commitment to data protection.
Lumo 2.0 now incorporates advanced visual processing capabilities through image analysis and generation tools. Users can upload images directly into the interface for contextual analysis or modification, while the chatbot can also create images from text prompts. This functionality competes with established multimodal AI features in popular assistants like Gemini and ChatGPT through enhanced contextual awareness while preserving privacy.
The update dramatically improves Lumo’s core functionality with Project integration. Users can now maintain persistent memory settings across Proton’s ecosystem – spanning email, cloud storage, and workflows – with all preferences securely stored under user control. The system learns from interactions without compromising data security through zero-access encryption architecture.
Technical improvements deliver a 76% faster response rate compared to the previous version, according to internal benchmarks. A “thinking mode” enables deeper problem-solving capabilities for complex queries through multi-step reasoning processing.
“Lumo 2.0 represents a complete architectural redesign,” stated Andy Yen, Proton’s founder and CEO. “Our innovation proves that users can access powerful AI capabilities without sacrificing privacy expectations.”
Available today through Proton’s free and premium subscription models (Plus and Professional tiers), Lumo maintains competitive performance with commercial AI services. The company emphasizes its unique position through cryptographic security measures including end-to-end encryption for all interactions and no server-side logging of conversations.
Proton reiterates that customer data remains exclusively accessible to users, with no repurposing for training datasets or third-party sharing. This privacy-first approach positions Lumo as an alternative for users concerned about data handling in mainstream AI platforms.
