As we have seen it happening in the last couple of years, every few months we get to hear about more job cuts. AI is always at the centre of it. This week again jobs and AI were the topics of discussion in the tech world. Thanks to Microsoft. The company announced an extraordinary voluntary exit offer, targeting nearly 60,000 employees across departments. Software engineers received emails asking them to consider voluntary retirement. And so did employees in departments like HR and Legal. Nearly 27 per cent of the entire Microsoft workforce has been put on a — soft — notice.
This is just a voluntary retirement scheme. But it is also Microsoft telling over one fourth of its employees they are no longer needed. Probably because AI can do their job.
Just a couple of days later, CEO of Proton, a company that makes privacy-focussed tools like VPNs and secure email service, said 2026 would be the last year he would hire an engineer. Apparently AI — largely Anthropic Claude — is generating almost 90 per cent software code within Proton. They don’t need software engineers anymore. Instead the plan, as reported by 404 Media, is to give AI tools to product managers and let them handle entire work from A to Z.
The overall crux of the matter seems to be that AI agents, driven by tools like ChatGPT and Claude, are now so good that many office roles and work can be done autonomously. The companies still need people, but fewer, and mostly to manage AI agents.
Sridhar Vembu, who nowadays misses no opportunity to weigh in on the talking point of the day, agrees. On Thursday he wrote on X: “Any job performed on the computer has this risk now: one smart person using AI can perform the work of 10 people. The risk is getting ever more real. The only thing we can do is to accept that it is coming and be prepared for it.”
So, is AI taking the jobs of people? Different people will answer differently. Data is all over the place, giving different conclusions depending on how it has been massaged and cajoled. There are areas where job loss is apparent, such as customer care. Can you recall when you last spoke to a human on a support line at apps like Amazon, Swiggy and Zomato? Then there are areas, such as cyber security, where companies might be hiring more. If you look at the departments that have been impacted by Microsoft’s voluntary exit offers, you will notice cybersecurity teams have not been touched.
But to see whether AI is impacting jobs or not, ignore data for now. Look at the job market and talk to people who are trying to find a job. It is a very tough scene out there. Even if the number of explicit layoffs is down relative to how it was a year ago, jobs aren’t available. Companies are quietly letting people go and are no longer hiring. A lot of companies are no longer adding new jobs. A lot of companies have replaced freelancers and outside consultants with AI agents, particularly in sectors such as graphics designing, communications and writing. These people, who were anyway on the margins of the formal job sector, are no longer finding work.
AI is changing knowledge work at a fundamental level. Consequently, it is also changing the social contract we humans have with society as far as work is concerned. What sort of future will this change bring? I am not sure. Nobody is. Although Francis End-Of-History Fukuyama recently wrote AI will replace human labour in “white collar work that involves symbolic manipulation” that is to say “people sitting behind computer screens all day.” Essentially, one thing is certain: people will have fewer opportunities to do 9-5 jobs in an office. Will it be a good thing or bad? That depends on how our governments, and we the people, navigate the new realities.