Proton plans to re-interview some employees about how they use AI, and their answers could matter to their jobs. The company behind Proton Mail and other privacy-focused products says the resulting assessments will feed a wider talent review that may change roles, responsibilities or continued employment.
404 Media’s Joseph Cox first reported the plan on October 26, citing an internal email he obtained. Proton later posted what it described as the full email sent to some staff, confirming the interviews and their potential employment consequences. The evidence establishes an announced review, not completed dismissals.
According to Proton’s posted email, interviews are scheduled for November 4 to 11, after CEO Andy Yen interviewed more than 40 senior Engineering and Product leaders. The next round will use a pool of leaders who have aligned on assessment criteria and who work outside each interviewee’s team and reporting line.
An interviewer outside the employee’s team and reporting line documents an assessment. That assessment informs a broader organizational and talent review, which may change roles, responsibilities or continued employment. Proton’s email describes no automatic pass-or-dismiss test and does not identify who makes final decisions.
Not everyone will be called. Proton says invitations will be prioritized according to its organizational assessment needs, and that receiving one, or not receiving one, is not itself an employment outcome. The company has not said how many workers are covered or which departments, countries or employment categories are included.
Employees who are invited must offer more than a list of AI tools. Proton’s prompts ask for concrete examples, what changed in the employee’s work, what succeeded or failed, what the employee learned and would do differently, and how the role could change. The public material provides no scoring weights, thresholds or standards tailored to particular jobs.
Proton also says sound AI use requires knowing when to avoid the technology and weighing privacy, security, accuracy and actual value. That distinction matters at a company built around privacy. Keeping sensitive information out of an AI tool could itself demonstrate good judgment, but the published prompts do not explain how interviewers will assess work for which AI is irrelevant or unsafe.
Proton denies using AI as a pretext for layoffs. It says it has a talent shortage, is hiring and wants faster product delivery, stronger future leaders and a better understanding of the support employees need. Those are the company’s stated reasons. They do not rule out individual employment endings, which its email expressly contemplates.
Another employer has already encountered the risk of rewarding AI activity instead of useful work. In a Fast Company interview, Duolingo CEO Luis von Ahn said the company removed a blanket AI-use evaluation after employees began using the technology for its own sake. He said performance reviews should measure contribution and acknowledged that AI may not help every role or project. Duolingo’s experience does not predict Proton’s results, but it shows the incentive a poorly designed assessment can create.
Proton has made the interviews consequential without revealing how consequential they will be. Its email does not say how much weight an assessment carries, what other performance evidence counts, or whether employees can improve or challenge a result before their jobs are affected. That missing decision process will determine whether Proton rewards the judgment it says it wants.