A new Microsoft report based on an analysis of 200,000 real conversations with AI tools such as ChatGPT, Gemini and Copilot identifies which professions are most exposed to AI-driven transformation. The findings are surprising: the biggest changes lie ahead not for manual workers but for office employees and knowledge specialists.

For the Polish labour market, where millions are employed in office sectors, this is a signal to act immediately.

A methodology that changes the rules of the game


Previous reports on AI's impact on the labour market relied mainly on theoretical models and surveys. Microsoft decided to take a different approach – and it is precisely this decision that makes the new study unique.

Instead of speculating, researchers analysed 200,000 genuine user conversations with the Bing Copilot tool conducted over nine months of 2024 in professional settings.

Based on this data, an indicator called the AI applicability score was created – a number between 0 and 1 that measures how often and how effectively artificial intelligence is used to perform specific professional tasks. The indicator takes into account both the frequency of use and the extent to which AI actually helps – not merely whether it is available.

This approach focuses on the real impact of technology rather than hypothetical scenarios.

"We are not asking whether AI could perform a given task. We are asking whether employees actually use it for that purpose and whether it brings measurable results. This is a fundamental difference in the way we think about labour market transformation." – Microsoft Research team, AI Applicability Study 2024 report.


The results were mapped onto the O*NET occupation classification – the most comprehensive database describing the activities performed within thousands of professions. This made it possible to precisely determine which tasks and professions are genuinely being transformed by generative AI tools.


Occupations with the highest AI applicability score


The study's findings are clear: artificial intelligence has the strongest impact on work based on knowledge, communication and information processing.

At the top of the list are three broad occupational categories: computer and mathematical specialists, office administration workers, and sales specialists dealing with information delivery and customer service.

This is no coincidence. All these professions share one common denominator: their daily work is filled with writing, data analysis, email communication and document processing – and these are exactly the tasks that tools such as ChatGPT, Microsoft Copilot and Google Gemini perform with growing effectiveness.

The study found that in the document management group, AI allows employees to process up to 10% more documents in the same amount of time.



  • IT specialists and data analysts – writing code, documentation, log analysis, creating technical reports

  • Administrative staff and assistants – handling correspondence, managing calendars, preparing presentations, taking minutes

  • Salespeople and customer service specialists – preparing offers, responding to enquiries, analysing customer needs

  • Financial analysts and accountants – processing data, preparing reports, forecasting

  • Marketing and communications specialists – creating content, analysing campaigns, editing materials


What does this mean for the office worker?


The key finding of the Microsoft report is: AI does not replace employees but takes over the routine elements of their work. In practice, this means that a marketing specialist, instead of spending three hours preparing a draft report, will delegate the task to Copilot and devote that time to strategic decisions and creative work.

Similarly, a financial analyst, instead of manually aggregating data from a dozen spreadsheets, will use automation and focus on interpreting the results.

"Artificial intelligence shifts the centre of gravity of human work towards situational judgement, oversight and interpersonal contact. This is not replacing the human – it is the evolution of their role." – Satya Nadella, CEO of Microsoft.


It sounds optimistic – and to a large extent this is consistent with reality. But the same mechanism carries a serious question: if one employee equipped with AI can do the work that three used to do, what happens to the other two? The report deliberately avoids forecasts about layoffs, yet the implications are clear to any manager.

AI-resistant professions – who has reason to be calm?


The lowest AI applicability scores are recorded by professions requiring physical presence, manual precision or complex interactions with an unrepeatable environment. Construction, transport, catering, agriculture – these sectors largely remain beyond the reach of current generative AI capabilities.

Bricklayers, roofers, car mechanics, construction machine operators – their work takes place in a different environment every day, requires adaptation to changing conditions and solving problems that are hard to catalogue in the form of a prompt. ChatGPT will not pour foundations or replace an engine's timing belt.

Microsoft's AI report: which professions is artificial intelligence transforming?



  • Crafts and manual work: bricklayers, roofers, mechanics, welders, electricians – no susceptibility to generative automation

  • Healthcare and education: nurses, primary school teachers, carers – growing demand for empathy and relationships

  • Agriculture, forestry and physical transport: truck drivers (until autonomisation), farmers, forestry workers

  • Surgeons and dentists: the physical dimension of work still limits AI's application to a diagnostic support role


Data for Poland confirms this picture.

The Polish Economic Institute indicates that the lowest exposure to AI automation concerns agriculture, construction and catering – sectors employing a total of around 1.66 million people. That is almost 10% of all employment in Poland.

Context of the Polish labour market


The Polish economy faces a particular challenge.

On the one hand, the business services sector (BPO/SSC) is growing dynamically, employing tens of thousands of specialists performing exactly the tasks that AI takes over most willingly: customer service, document processing, financial analysis. On the other – there is a shortage of qualified workers in construction and industry, where AI has limited reach.


"Poland is in a unique position: we have a strong knowledge services sector that is directly at the centre of AI transformation, and at the same time huge staffing shortages in professions that AI will not touch for a decade. The problem is not a lack of jobs – the problem is a mismatch of competences." – Piotr Arak, Director of the Polish Economic Institute.


Forecasts say that by 2027 as many as 75% of professions in Poland will retain low susceptibility to full automation by AI. However, this does not mean no change at all – it means that the transformation will be gradual and will primarily affect how work is performed, not whether it exists at all. Adaptability, critical thinking and collaboration with AI tools will become crucial.


New professions that AI creates


Every technological revolution not only destroys old jobs but creates new ones. The AI revolution is no exception. Experts point to a whole spectrum of new roles that are only just entering the market or will become a mass phenomenon in the coming years.

AI ethics specialists – people ensuring that artificial intelligence systems operate in line with values and legal regulations, including the EU's AI Act. Prompt engineers – experts in formulating queries to language models in a way that maximises their usefulness. AI trainers – specialists evaluating and correcting AI model outputs. Finally, data storytellers – analysts capable of turning raw data into narratives understandable to boards and decision-makers.

The Future of Jobs 2025 report prepared by the World Economic Forum forecasts that by 2030 a global total of 170 million new jobs linked to technology and digital transformation will appear, while around 92 million positions performing routine tasks will disappear. The balance is positive – but only for those who manage to prepare in time.

What does this mean for marketing and communications?


The marketing industry finds itself at a particularly interesting point in this transformation. On the one hand, AI tools (Jasper, Copy.ai, Midjourney, DALL-E) are revolutionising content production, data analysis and campaign personalisation. On the other – strategic thinking, creativity and understanding human emotions remain the human domain.

For the Polish marketing market, this means an accelerated change in the working model. Implementing Copilot or other GenAI tools in an agency or marketing department does not eliminate the need for experienced strategists and creative directors – quite the opposite, it raises the bar.

When AI handles routine copywriting and basic analytics, more is demanded of humans: a deeper understanding of the brand, courage in decision-making and the ability to build authentic relationships.

How to prepare for AI transformation?


Experts agree: passively waiting for events to unfold is the worst possible approach.

Both employees and companies should already take concrete adaptive action.


  • Audit your own professional tasks – which activities do you perform daily and which of them could AI do?

    This is the starting point for an adaptation strategy


  • Invest in soft skills – communication, empathy, leadership and critical thinking are competences that AI will not replace in the foreseeable future

  • Learn to use AI tools – this is not about programming but about effectively using Copilot, ChatGPT or industry tools as work assistants

  • Develop analytical competences – the ability to interpret data and draw conclusions will become increasingly important as AI takes over its processing

  • Build a network of relationships and expert positioning – industry knowledge and interpersonal relationships are resources that an algorithm does not possess


Microsoft Corporation

Microsoft is one of the world's largest technology corporations, founded in 1975 by Bill Gates and Paul Allen. The company employs over 220,000 people globally and generates annual revenues exceeding £190 billion. In the field of AI, Microsoft is a key investor in and partner of OpenAI, and tools from the Copilot family are built into almost the entire product ecosystem – from Windows through Microsoft 365 to the Azure cloud platform. A fun fact: Microsoft's investment in OpenAI has reached a value of over £10 billion, making it one of the largest bets on AI in the history of technology.

Read also:

Company information

Microsoft Corporation is one of the world's largest technology corporations, founded in 1975 by Bill Gates and Paul Allen. The company employs over 220,000 people globally and generates annual revenues exceeding £190 billion.

In the field of AI, Microsoft is a key investor in and partner of OpenAI, and tools from the Copilot family are built into almost the entire product ecosystem – from Windows through Microsoft 365 to the Azure cloud platform.

A fun fact: Microsoft's investment in OpenAI has reached a value of over £10 billion, making it one of the largest bets on AI in the history of technology.