Proton CEO Andy Yen has cancelled a round of nearly company-wide re-interviews in which staff were to explain how they use AI in their work. The step followed a report by the US outlet 404 Media, which the Swiss company's staff had talked to, and came one day after that report, according to IT portal it-daily.
Six employees told the outlet that they feared losing their jobs if they used AI too little. According to 404 Media's sources, Yen had urged developers to rely more on AI instead of writing every line of code by hand. A staff proposal to run penetration tests on newly written code before release was reportedly dismissed by Yen in a security channel in late August. Staff said a culture had developed in which security concerns were rarely raised.
In a recording of a roughly two-hour staff meeting, which 404 Media obtained, Yen said the decision to cancel the re-interviews had been taken after talks with the leadership team and the board. The AI criteria are now to be part of regular performance reviews. Yen said no jobs would be cut because of AI. Proton did not respond to 404 Media's request for comment for that report.
Users reacted with criticism. On Reddit many announced that they would end their paid Proton plans, and some called for Yen to step down. Proton Mail, Proton VPN and Proton Drive are all part of the company's offer. How many users actually cancelled is not known.
Proton later addressed the criticism in a public statement on 30 October. The company apologized and said its use of AI would never come at the expense of security. It denied installing tracking software on staff laptops to measure how much AI-generated code employees produced, describing that claim as a misunderstanding about device-management software.
Proton said it had added layers of security review over the previous year in response to hostile AI tools scanning its open-source code for vulnerabilities. It promised to publish its internal AI policies, explain its code-review and security-review procedures, and release several internal tools as open source, including more capable versions of its Lumo AI assistant. PCWorld reported on 9 November that Yen had also argued on Reddit that the spread of AI tools meant Proton's code was receiving more third-party scrutiny than before.
On 1 November, Proton CTO Bart Butler published the AI manifesto that had previously circulated internally. It proposes that engineers increasingly manage AI models and agents and validate their output rather than writing all code themselves. Butler said engineers remain responsible for the results and stressed the importance of testing, sound architecture and expertise, while describing the document as a general direction rather than a detailed security rulebook.
Proton has been majority-owned by the non-profit Proton Foundation since June 2024. Its AI assistant Lumo stores chat histories encrypted and, according to Proton, does not use them to train AI models.