Lumo: Proton’s AI Assistant review

I wanted to do a review of Lumo, an AI Assistant by Proton. First you might be asking, who is Proton? Well they are a privacy focused software and service company based out of Switzerland. Some of their most popular services are Proton Mail (email), Proton VPN (Virtual Private Network), and Proton Drive (end to end encryption file storage).

Lumo prides itself on privacy such as: no logs of questions and responses, zero access encryption (only you have access), and your data is in your control. Also their source code is open source and available here. https://github.com/ProtonMail/WebClients/tree/main/applications/lumo/ When doing some research, it seems like not all of its open source, but I didn’t get time to dig into the specifics, so use caution.

General information and grammar

I just finished up my post about Hoh Rain Forest, so I decieded to ask it

tell me about hoh rain forest

Results

The H oh Rain Forest is a lush, temperate rainforest located in the western foothills of the Olympic Mountains within Olympic National Park in Washington State, USA. <plus more technical details>

First thing I noticed was the spelling error in Hoh, with an extra space. I then decided to return that same sentence back to the chatbot and asked it if there was any mistakes?

Is there any mistakes in this - The H oh Rain Forest is a lush, temperate rainforest located in the western foothills of the Olympic Mountains within Olympic National Park in Washington State, USA.

Response from Lumo:

Your sentence is clear and mostly correct, but there are a couple of small tweaks you could make for optimal grammar and style:

Original:
The H oh Rain Forest is a lush, temperate rainforest located in the western foothills of the Olympic Mountains within Olympic National Park in Washington State, USA.

Suggested revision:
The H oh Rain Forest is a lush, temperate rainforest located in the western foothills of the Olympic Mountains, within Olympic National Park in Washington State, USA.

Why the change?

Adding a comma after “Olympic Mountains” creates a natural pause before the phrase “within Olympic National Park.” This improves readability by separating the two geographic descriptors.
Everything else—spelling, capitalization, and overall structure—is spot‑on. Great job!

Didn’t even catch the spelling error. I wanted to see what chatgpt would say and here is what it returned

Yes, there’s a small formatting issue in your sentence. You wrote “H oh” with a narrow space between the “H” and “oh,” which is not standard. The correct name is “Hoh Rain Forest” (no space or special character).

A slightly polished version of your sentence could be:

“The Hoh Rain Forest is a lush, temperate rainforest located in the western foothills of the Olympic Mountains within Olympic National Park in Washington State, USA.” ✅

Grammatically and factually, the rest of the sentence is fine.

If you want, I can also suggest a slightly snappier, SEO-friendly version for a blog or guide.

Rewriting for readability

In a previous post I went and asked some LLM some basic questions and ranked them based on my opinions on what I thought was the best responses. I decided to revisit and reuse some of those questions to evaluate Lumo.

Rewrite this for better readability – the lazy yellow dog was caught by the slow red fox as he lay sleeping in the sun

Response

The sleepy yellow dog was snared by the sluggish red fox while it rested in the sunshine.

Comparing this to the answers I got from my LLM test, I feel like this response really deviated from the originally intent.

Next I asked Lumo

Tell me the ethical concerns related to AI

Response

Key Ethical Concerns Around Artificial Intelligence
Artificial intelligence brings powerful capabilities, but its deployment raises a suite of ethical challenges that societies, developers, and policymakers must grapple with. Below are the most frequently discussed concerns, along with brief reflections on why each matters and what tensions they create.

<Removed large portions of the post to not blot post>

1. Bias and Fairness
2. Transparency and Explainability
3. Accountability and Responsibility
4. Privacy and Data Governance
5. Job Displacement and Economic Impact
6. Security and Robustness
7. Misuse and Dual‑Use Risks
8. Autonomy and Human Agency
9. Environmental Footprint
10. Governance and Global Equity

Overall, I thought this last response was pretty good, contained tons of detail, and was very thorough.

While this was only a short test of Lumo, the first post had a spelling error and the second struggled with a simple rewrite. There is plenty of room for improvement, but I still think it has a lot of potential, especially given the privacy concerns around AI. I did appreciate that the responses are much faster then using a local LLM (eg through Ollama). Overall, Lumo shows promise and could become a strong option as it continues to mature and improve.