Proton says it will not re-interview staff about their AI use or make AI use a reason for job losses, following a backlash over its push to use AI in building privacy products. The company has also promised more public detail on its tools and security reviews, putting its assurances to the test.
Watch Desk analysis
What happened
Proton’s commitments include a promise that using AI will not come at the expense of security and that it does not track employees’ laptops to measure how much AI-generated code they produce. The company also says it will open-source several internal tools and publish information about its code and security reviews.
The response follows the publication of an internal AI engineering manifesto by Proton CTO Bart Butler. As described by TechRadar, Butler’s memo urges engineers to focus on directing AI models and agents rather than writing code themselves. TechRadar says Proton also plans to make more capable versions of its internal Lumo AI available outside the company.
Why it matters
Proton sells privacy-focused products, so the question is not simply whether its engineers use AI. It is whether the company can explain how that use fits with its security commitments, and give staff and customers evidence they can scrutinise. Open-sourcing tools and publishing review practices would offer more to assess than a promise alone.
The employee commitments matter too: Proton says AI use will not trigger job losses, even as its manifesto describes a substantial change in what it expects engineers to do. Those are distinct claims, and the company’s follow-through will matter more than the wording of either.
Our read
This is a meaningful response, but the lasting value depends on what Proton publishes. More capable Lumo tools may interest customers; clear security practices and verifiable limits on workplace monitoring are the more important test. Privacy is a fine promise. Showing your workings is better.
What to watch
- Which internal tools Proton releases, and under what terms.
- What it publishes about code review and security review practices.
- Whether the promised Lumo capabilities become available outside the company.
Discussion spark: Should a privacy-focused company have to publish its AI-related security practices before asking customers to trust new AI features?
Sources and evidence
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.