Proton Users Cancel Subscriptions After CEO Reportedly Called Security Caution 'Loser Thinking'

What happened: from internal messages to public controversy

Decrypt reports that Proton users say they have canceled subscriptions after CEO Andy Yen was reported to have pressured engineers to embrace AI and dismissed calls for more security testing as "loser thinking." Everything below comes from Decrypt's report, some of it relayed from the news outlet 404 Media.

According to Decrypt, citing 404 Media, an employee in August asked for a few days to run penetration testing on new code, meaning experts try to break it before release. Yen replied: "That's bullshit and loser thinking. Of course it's possible to be fast and secure." Decrypt notes that 404 Media's investigation drew on nine current and former employees, internal messages and meeting recordings.

According to Decrypt, citing 404 Media, Yen also urged employees to stop hand-writing code and use AI more, and mentioned re-interviewing staff to understand whether and how they use AI in their work.

Proton's response and engineering manifesto

According to Decrypt, Yen canceled the re-interviews on October 30. The assessment would have asked employees to explain how they use AI, and staff feared their answers could determine whether they kept their jobs. Proton also acknowledged to 404 Media that management had sometimes communicated too bluntly and that employees were anxious.

According to Decrypt, CTO Bart Butler published an engineering manifesto on November 1 arguing that developers should direct large language models (LLMs, AI systems that generate text and code) rather than treat them as occasional assistants. The manifesto also insists that expertise and testing still matter, and that engineers remain responsible for work models cannot reliably handle. Decrypt quotes the manifesto: "Your job is to help the LLM write better code faster."

Decrypt also reports that the manifesto calls on the whole company to invest in automation and treats avoiding AI as an "existential threat" to the company's mission succeeding.

What this means for ordinary users

Decrypt notes that the controversy hits Proton's core selling point: users trust it to hold information they want kept private. Some customers are questioning whether a faster release pace will come at the expense of their security.

Decrypt also points out that AI tools can themselves bring security problems. For example, Microsoft researchers said a patched Claude vulnerability could leak secret access credentials through malicious instructions in data read by a coding assistant; meanwhile, Google's PageBreak system can search its own applications for exploitable flaws and check whether attacks actually work. In other words, AI can both create risk and help find weaknesses.

Browse the latest news in this topic

• Crypto & blockchain