Proton has introduced Lumo 2.0, adding image generation, editing, image analysis, web search, memory and reasoning modes to its privacy-focused AI assistant. Proton says uploaded and generated images are protected with zero-access encryption.
Proton has released Lumo 2.0, an update to its privacy-focused AI assistant that adds image generation, image editing and image analysis alongside new reasoning, web-search and memory features.
The release broadens Lumo beyond a text chatbot. Users can generate images from written prompts, upload images for analysis, and make edits through natural-language instructions, according to Proton’s product announcement. The company positions the additions as part of an effort to offer modern AI functions while preserving the privacy standards associated with its encrypted email, storage and VPN products.
Proton says Lumo’s generated and uploaded images are protected by zero-access encryption. In practical terms, that is the company’s claim that content is encrypted in a way intended to prevent Proton itself from accessing users’ data.
That design is an important distinction in a market where AI image tools commonly require users to submit prompts and media to a provider’s cloud infrastructure. Proton’s announcement presents privacy as a core product feature rather than an optional setting, though users should still review the service’s current terms and product documentation to understand how individual features, including web search, handle data.
Lumo 2.0 also introduces reasoning modes intended for more complex requests, according to Proton. The update adds web search, enabling the assistant to retrieve information from online sources, and memory features that can retain useful context between conversations.
TechCrunch likewise reported that the release includes image recognition, prompt-based image generation, persistent memory and a thinking mode. Together, those functions bring Lumo closer to the broader set of capabilities offered by mainstream AI assistants, while Proton continues to differentiate the product through its privacy-focused positioning.
Proton’s product updates page describes Lumo as part of the company’s growing portfolio of privacy-oriented services. With version 2.0, the assistant is being positioned for tasks that range from drafting and research to visual creation and image understanding.
The expansion also illustrates a broader tension in consumer AI: more capable assistants tend to need access to more context, files and online information, while privacy-conscious users want tighter limits on who can see that material. Proton’s answer is to combine multimodal AI features with its stated encryption model.
For users considering Lumo 2.0, the most consequential changes are likely to be the arrival of image tools and memory. Those features can make an assistant more useful over time, but they also make transparency around storage, retention and data access especially important. Proton’s release frames zero-access encryption as the safeguard underpinning that approach.
Sources: Proton’s Lumo 2.0 announcement; Proton Lumo product updates; TechCrunch coverage.
Users can generate images from written prompts, upload images for analysis, and make edits through natural language instructions, according to Proton’s product announcement.
The company positions the additions as part of an effort to offer modern AI functions while preserving the privacy standards associated with its encrypted email, storage and VPN products.
Privacy is central to Proton’s pitch Proton says Lumo’s generated and uploaded images are protected by zero access encryption.
Continue reading
