Proton pledges to open-source internal AI tools after backlash
After reports it would re-interview staff about AI use, Proton pledged to open-source internal AI tools and publish its AI use policy. Write your AI policy first.
Proton, the company behind Proton Mail, says it will open-source internal AI tools and publish where staff may and may not use AI. The pledge came on October 30, after a week of bad press about how it planned to push AI on its own employees. The lesson for a small business is simple: if you don't write your AI policy down, someone else will describe it for you.
What actually happened
On October 26, 404 Media reported that Proton told staff it would "re-interview" them about how they use AI at work. According to the internal email 404 Media obtained, the interviews could affect continued employment.
Proton then posted a statement on X on October 30. As CyberInsider reports, Proton said it will:
- Release and open-source internal AI tools, including more capable versions of its Lumo assistant.
- Publish where AI is allowed and banned inside the company, and why.
- Publish how it does code and security review.
Proton also denied that it re-interviews staff about AI use, denied that AI interviews will cost anyone their job, and said reports of laptop tracking came from a misunderstanding of device management software. Per CyberInsider, Proton gave no release date, no license, and no names for the tools beyond Lumo. We could not read Proton's original post directly, so we rely on CyberInsider's account of it. Treat the pledge as a promise until code ships.
Why an AI use policy matters for your business
Customers now ask how your work was made. Proton's pitch is that you can check how its code was produced, not only the code. Your customers won't audit your repo, but they will ask if AI wrote their quote, their ad copy or their support reply. Have one honest page that answers.
Staff need the rules before the mandate. Proton's trouble started with an internal email about AI and job security. If you want your team to use AI, publish what is allowed first: which tools, which data never goes in, who reviews the output. Then measure. In that order.
A one-page policy beats a pledge. You don't need open-source tools. You need a page with three lists: allowed, banned, needs review. Put customer data, credentials and contracts in the second list. Update it every quarter.
Key takeaways
- 404 Media reported on October 26 that Proton planned to re-interview staff about AI use
- On October 30 Proton pledged to open-source internal AI tools, including more capable Lumo versions
- Proton also pledged to publish where AI is allowed and banned internally, and how it reviews code
- No release date, license or full tool list was given; treat it as a promise until code ships
- Write a one-page AI policy (allowed, banned, needs review) before you push AI on your team
Rolling out AI to your team without the drama? We set up AI tools with written use rules, data boundaries and review steps built in, so your staff know the rules before anyone asks. Talk to us about your rollout.
Sources: 404 Media, CyberInsider.
- #ai-policy
- #proton
- #open-source
- #ai-adoption
- #trust
Tommy Rush — Founder, Rush Commerce
Operator turned builder. 15+ years running operations — now shipping the systems businesses run on. More
Get The Rush Report weekly — one email, zero fluff.
Keep reading
Thrive Holdings Buys ADT for $15.3B: AI Rollups Hit Home Services
Thrive Holdings will buy ADT for $15.3B, taking its tech-first rollup into home security. What an AI rollup in home services means for local operators.
Read itDelaware AI company bill: agents run the business, Comer probes
Delaware's draft AI company bill would let an AI agent run a business day to day. Congress is now asking questions. Liability still lands on a human.
Read it