Microsoft’s new study reveals which jobs AI will kill first

New Delhi: A new study by Microsoft Research, conducted in collaboration with OpenAI and LinkedIn, has highlighted the growing impact of generative AI on the workforce, identifying 40 jobs most vulnerable to disruption and 40 that remain largely shielded, for now.

Drawing from over 200,000 real-world interactions with Microsoft’s Copilot AI tool, the study offers a data-backed look at how AI is already reshaping roles across industries. Notably, professions linked to advertising, writing, and customer service are among the most exposed to AI-led automation.

Leading the list of jobs most susceptible to AI disruption are interpreters and translators, historians, passenger attendants, and service sales representatives. Also flagged were writers and authors, customer service executives, CNC tool programmers, telephone operators, and roles like radio hosts, editors, political scientists, and public relations specialists.

These professions score high on what the study terms “AI applicability”—meaning the tasks they involve significantly overlap with AI capabilities such as language generation, data analysis, and task automation. The result is not necessarily job elimination, but erosion of core responsibilities, making it essential for workers to upskill or adapt.

Advertising roles in the crosshairs

The advertising industry, in particular, is facing accelerated disruption. AI is rapidly taking over campaign execution tasks such as targeting, content creation, trend forecasting, and client outreach. Among the most at-risk roles:

• Advertising sales agents: AI can now automate outreach, optimise placements, and even negotiate ad buys.
  
  
• Market research analysts: Tools can perform large-scale data analysis and consumer behaviour modelling with greater speed.
  
  
• Writers and authors: Generative AI is capable of creating everything from ad copy to video scripts, reducing reliance on human writers.
  
  
• PR specialists: AI tools are increasingly able to draft releases, manage media responses, and monitor sentiment.
  
  

The study emphasises that the shift is not a total replacement but a significant reduction in demand for certain skills, particularly those involving routine or repeatable tasks.

Jobs least affected by AI

On the other end of the spectrum, roles that involve manual labour, physical dexterity, or emotional intelligence remain largely unaffected. These include professions such as:

• Construction workers, roofers, and electricians
  
  
• Cleaners, cooks, and dishwashers
  
  
• Healthcare workers like nursing assistants, surgical aides, and massage therapists
  
  
• Infrastructure operators, such as dredge operators and cement masons
  
  

These jobs require physical presence, human judgement, or context-specific action—areas where AI tools and robotics still lag significantly.

The findings come at a time when tech leaders have begun issuing cautionary signals. OpenAI CEO Sam Altman has warned that entire categories of white-collar jobs, especially in customer service, may vanish as AI becomes more capable. 

Similarly, Anthropic CEO Dario Amodei has projected that up to 50% of entry-level white-collar jobs could disappear in the next five years, potentially pushing unemployment rates to 10–20%, unless governments and industries proactively manage the transition.

Microsoft, too, is undergoing its own transformation. Even as it lays off nearly 15,000 employees, it continues to pour resources into building AI infrastructure, such as Azure data centres powering Copilot and ChatGPT.

Ultimately, while AI may be automating more tasks, the human edge still lies in adaptability, emotional insight, and the ability to craft stories that resonate. For now, the most future-ready workers will be those who learn to work with AI, not against it.