www.it-connect.fr
Proton’s AI Push Sparks Internal Backlash as CEO Dismisses Security Testing as “Loser Thinking”
In October 2026, 90% of the code for Proton’s tools was reportedly written by AI. Beyond that figure, there is growing concern among several employees at the Swiss company and pressure from leadership, starting with CEO Andy Yen. Here’s what 404 Media’s investigation reveals.
For context, Proton is the Swiss company behind Proton Mail, launched in 2014 after the Edward Snowden revelations. Since then, the Proton ecosystem has expanded significantly with many tools and services, including Proton Pass, Proton VPN, and more. Since July 2025, it has also included Lumo, Proton’s AI designed with privacy in mind.
In public, Andy Yen, founder and CEO of Proton, takes a cautious stance on AI, and at times even a reluctant one. In an interview with WIRED in August, he described AI as a form of surveillance and acknowledged that Proton had entered the field reluctantly. Even so, he believes privacy-focused companies also had to move into this area, so the future would not be left entirely to tech giants.
Internally, the tone would reportedly be very different. On October 29, 2026, Joseph Cox of 404 Media published a lengthy investigation into Proton. It is based on testimony from eight employees, one former Proton employee, internal Slack conversations, meeting recordings, and an internal memo written by CTO Bart Butler.
Buckle up, this gets intense.
Security testing dismissed as “loser thinking”
At the end of August, in the Slack channel dedicated to security, an employee asked to take a few days to perform a penetration test on freshly written code before it went into production. "That's bullshit and loser thinking. Of course we can be fast and secure." Andy Yen replied, according to messages reviewed by 404 Media. He then urged teams to stop writing code by hand and rely more heavily on AI. That prompted another engineer to post a message saying security seemed to be a low priority these days... without getting a direct response. Not exactly reassuring.
According to 404 Media, Proton has also put together a real internal action plan to promote AI:
- A dedicated email address: employees can use it to report anything slowing AI adoption, so management can remove the blocker.
- Monitoring software: installed on workstations, it tracks how much code is generated by AI.
- The end of engineer hiring: Andy Yen announced that this would be the last year Proton hires engineers, with the company then aiming for profiles focused on building products with AI.
Then, on a Friday afternoon in September, Bart Butler posted an internal memo on Slack titled “AI @ Proton.” "Your job has fundamentally changed. It is no longer about writing code.", the document says. If Proton does not make this shift, it would be an "existential threat to the success of the mission itself", the CTO warns. He also wants, in the long term, code that is mostly reviewed by AI rather than by humans. In a meeting, he reportedly said that 90% of the code produced in October had been written by AI, largely with Anthropic’s Claude, according to several employees.
Another episode came in May, during the deployment of post-quantum encryption in Proton Mail. Unhappy that the communication highlighted a limitation (older emails were not covered), Andy Yen reportedly asked a team member to correct the record on Reddit using his personal account.
A string of interviews that feels heavy-handed
On October 26, Proton announced by email that some employees would be “re-interviewed” about their use of AI from November 4 to 11, 2026. Each affected employee will meet with a manager outside their team and reporting line, selected for their AI expertise. A simple list of tools will not be enough: concrete examples will be required to prove how AI is being used. Andy Yen had already questioned more than 40 leaders across product and engineering teams. Most notably, the email warns that these interviews may affect each person’s continued employment.
"We do not make layoffs and use AI as an excuse." Proton said in a comment posted on r/ProtonMail. The company pointed to a talent shortage and, despite the tense atmosphere internally, ongoing hiring.
When contacted by 404 Media, Proton did not dispute any part of the investigation. Its spokesperson acknowledged internal concerns and a management style that can sometimes be too harsh. The situation appears tense, and another internal document also indicates that Proton is not on track to reach its annual recurring revenue target.
If you use Proton, could these statements make you doubt or question the use of its services?
