Proton has pledged to release and open-source internal AI tools, including more capable versions of its Lumo assistant, and publish details of its software review processes following criticism of its AI strategy and workplace culture.
In a statement published yesterday on X, the company said it would disclose where AI is permitted internally, where it is prohibited, and the reasoning behind those decisions. The commitments accompanied an apology and responses to allegations about employee monitoring, security practices, and AI-related job assessments.
Proton, the privacy-focused provider behind Proton Mail, Proton VPN, Drive, and Pass, also operates Lumo, an AI assistant built around open-source language models. According to its privacy documentation, Lumo runs those models on Proton-controlled servers and protects saved chat histories with zero-access encryption.
The planned releases would expose tools that have so far remained inside the company, including versions of Lumo that Proton describes as having greater capabilities than those available publicly. However, the announcement does not specify those capabilities, identify the other tools, or provide a release schedule or licensing details.
Alongside the software, Proton promised to document how it conducts code reviews and security reviews, giving outsiders more visibility into the development practices behind its products.
“Instead of just being able to see the final open-source code that we produce to verify security, you will also be able to assess how that code was produced,” Proton stated.
Proton’s response to community backlash
The statement follows reporting by 404 Media’s Joseph Cox on October 26, based on an internal email outlining staff “re-interviews” about AI adoption. The email said assessments would inform an organizational review that could affect roles, responsibilities, and continued employment, with interviews scheduled for November 4–11.
In an October 30 follow-up, 404 Media reported that CEO Andy Yen had canceled the interviews, citing six current employees and a recording of an all-hands meeting. Proton’s subsequent public statement says it is not re-interviewing staff about AI usage and that there are no job losses due to AI re-interviews.
Proton also denied installing software on employees’ laptops to measure AI-generated code output, attributing that allegation to a misunderstanding of mobile device management software. It said it had introduced additional security review layers over the past year to address hostile AI systems scanning its open-source code for vulnerabilities.
CyberInsider contacted Proton for comment when the original 404 Media report appeared but received no response.
Leave a Reply