UpTrajectory Review

Proton, the Swiss firm behind Proton Mail, has told some employees they must sit for a second interview about their AI habits at work, and the outcome could decide who stays and who goes. Internal messages cited by 404 Media show chief executive Andy Yen pressing engineers to stop writing code by hand and lean on AI tools instead. The story surfaced through reporter Joseph Cox and reached a wider audience via The Next Web, which cut the piece short behind a continuation link. That truncation matters: the public record so far rests on leaked messages and one outlet's account, not on a formal Proton policy document or a verified company statement. Readers should treat the details as credible but incomplete until Proton responds on the record.

For a small-business operator, this is not a distant tech-industry drama. Proton built its brand on privacy, security, and trust; its customers pay for email and VPN services precisely because they believe the company exercises restraint. If Proton now ties employment to AI adoption, it signals that even privacy-first firms view AI fluency as a baseline competency, not a bonus skill. That shifts the burden onto every owner who hires developers, support staff, or operations people: the question is no longer whether your team uses AI, but whether they use it well enough to justify their salary. It also raises a harder question for firms that sell trust: can you demand aggressive AI adoption from staff while promising cautious, human judgment to customers?

What is genuinely new here is the mechanism. Firms have cut jobs and blamed AI before, but Proton appears to be auditing individual behavior and making employment contingent on the audit. That crosses from strategy into surveillance. We are skeptical of the framing that hand-written code is now a liability; plenty of secure, maintainable work still requires human reasoning, and Proton's own security reputation depends on it. At the same time, we agree with the underlying pressure Yen faces: AI coding tools have moved fast, and a firm that ignores them risks falling behind competitors who ship faster and cheaper. The tension is real, but the solution, re-interviewing staff under threat of dismissal, looks blunt and likely to breed fear rather than fluency.

The downstream effects cut unevenly. Senior engineers who hand-craft code for security reasons may feel punished for the very rigor that won Proton its reputation. Junior staff, who often learn by writing code manually, may skip that apprenticeship and produce work they cannot debug or defend. Customers who pay a premium for Swiss privacy may wonder whether AI-generated code introduces new vulnerabilities or weakens the human review that security demands. Other privacy-focused firms will watch closely: if Proton normalizes AI-linked job reviews, smaller rivals may copy the practice to cut costs, while some customers may defect to providers that promise slower, more deliberate development.

Watch for three things next. First, whether Proton issues a public clarification or doubles down; silence will let the leaked messages define the policy. Second, whether regulators or labor bodies in Switzerland take interest, since conditioning employment on tool adoption may raise workplace-law questions. Third, whether other privacy-first firms, from VPN providers to encrypted-messaging apps, follow suit or distance themselves. Operators should use this moment to audit their own AI expectations: write a clear internal policy, train staff on secure AI use, and decide what work must remain human. Do not wait for a leaked memo to force the conversation.

Takeaway: Set a clear, written AI-use policy now; do not wait for a leaked memo or a re-interview to define your team's expectations.