Proton to re-interview staff on AI, with jobs at stake, 404 Media reports

Chief executive Andy Yen called a request for extra security testing “loser thinking”, according to leaked Slack messages. Proton says it is not cutting jobs and is still hiring.


Proton to re-interview staff on AI, with jobs at stake, 404 Media reports

Andy Yen, founder and chief executive of Proton, at Web Summit in Lisbon in November 2021.

Image Credits Credit: Piaras Ó Mídheach / Web Summit via Sportsfile / CC BY 2.0 via Flickr (cropped)

Proton, the Swiss company behind Proton Mail, will “re-interview” some of its staff about how they use AI at work. The results may affect who keeps their job. Internal messages show chief executive Andy Yen telling engineers to stop hand-writing code and use AI instead, 404 Media reported.

Joseph Cox revealed the plan for 404 Media on Monday, from an internal email. On Thursday he followed it with an investigation. It draws on nine current and former employees, Slack chats, meeting recordings, and a manifesto by the chief technology officer.

Interviews that are “consequential”

The email calls the re-interviews “consequential”. Interviewers will record their assessments for a wider review. That review may lead to changes in roles and in “continued employment”, it says.

Yen began the process by interviewing more than 40 senior leaders in engineering and product. Not every employee will be called in. Each interviewer will come from outside the employee’s team, from a pool of leaders who “demonstrate strong AI capability”. The interviews run from 4 to 11 November.

Proton has since said on Reddit that it is not doing layoffs and using AI as an excuse, TechRadar reported. The company said it faces a talent shortage, is hiring, and wants to find out who needs more support.

What Yen wrote on Slack

In August, an employee asked for a few days to run security tests on new code before release. Yen called the idea “bullshit and loser thinking”, according to Slack messages seen by 404 Media. He told staff to stop handcrafting code and use AI more.

“This will be the last year that we hire engineers,” Yen wrote in the same channel, according to 404 Media.

After Proton announced post-quantum encryption in May, users on Reddit criticised a limitation: it did not cover older emails. Yen told the team that “perception is more important than truth”, 404 Media reported.

AI wrote 90% of Proton’s code in October, chief technology officer Bart Butler said in a meeting. 404 Media obtained the audio. Employees said they wrote much of it with Anthropic’s Claude. Proton also installed software on staff computers that tracks how much code AI writes.

Proton is not on track to hit its annual recurring revenue target this year, an internal document says. It blames “insufficient velocity”.

Proton’s response

A Proton spokesperson did not dispute 404 Media’s reporting, and told the outlet:

“In our determination to move fast, we have sometimes been too blunt or too slow in explaining ourselves to our people,” the spokesperson said.

In August, Yen told WIRED that privacy companies have to build their own AI. Otherwise, he said, they cede the future to companies that do not have users’ interests at heart. Proton’s chatbot, Lumo, does not log conversations or train on user data, the company says.

Amazon has warned its engineers against chasing AI usage metrics, and an Anthropic manager has dropped coding interviews. Microsoft this week offered buyouts to 60,000 staff.

Get the TNW newsletter

Get the most important tech news in your inbox each week.

Thanks! Check your inbox to confirm your subscription.

By subscribing you agree to our Privacy Statement.