AI Is Already Taking Jobs – Which Professions Will Disappear First

Research indicates that approximately 30% of jobs in Poland — more than 5 million positions — could be at least partially automated by generative AI. Office workers are particularly vulnerable: up to 71% of their tasks can be delegated to machines. Real-world examples from Poland already show AI displacing certain occupations: in banking, voicebots have taken over customer service (70 million conversations), while in logistics, warehouse robots are replacing manual workers.

Why AI Threatens Many Occupations

Artificial intelligence is no longer just a tool that supports workers — it is increasingly becoming their direct replacement. Robots, AGV systems and chatbots can perform repetitive tasks faster, cheaper and without breaks, making many positions redundant. Companies that can cut operational costs by 30–60% do not hesitate to invest in automation.

According to the latest data, 30.3% of positions in Poland — approximately 5.08 million jobs — are susceptible to automation by generative AI. These are not distant forecasts: some of these processes are already happening before our eyes. The OECD estimates that by 2030 as many as 85% of cashier positions could be automated. McKinsey points out that AI can already take over 71% of typical administrative tasks right now.

The impact of automation is not evenly distributed. Manufacturing and logistics feel it most strongly — physical robots and sorting systems replace people at assembly lines and in warehouses. Industries requiring empathy and complex judgement, such as education and medicine, are less exposed for now, although AI is beginning to make inroads there too — particularly in medical imaging and document management.

Poland’s Central Statistical Office (GUS) reports around 250,000 cashiers and around 250,000 accountants employed in the country. Both occupational groups are particularly at risk today — their tasks are largely based on routine operations that algorithms can perform with high accuracy and without fatigue.

Top 8 Professions in the AI Firing Line

The table below covers occupations where the risk of automation is highest — both due to the nature of the tasks performed and the availability of ready-made technological solutions.

Occupation	Reason for Automation	Risk (%)	Source
Cashier / sales assistant	Self-checkout, scan/image recognition	85% of tasks	OECD/WEP (2024)
Accountant / office worker	AI in ERP and OCR invoice processing, analytics	50% of tasks	McKinsey/WEF (2025)
Telemarketer / call centre agent	Chatbots and voicebots (AI handles 75% of calls)	70–75% of calls	WEF (2025)
Driver / courier	Autonomous vehicles and delivery drones	growing	V2X companies (2025)
Warehouse worker	Mobile robots (AGV) sorting and packing	50% of tasks	McKinsey/IFR (2024)
Receptionist / assistant	Booking systems and appointment chatbots	approx. 60–70%	BLS/WEF (2024)
Medical diagnostician (X-ray)	AI in medical imaging (DeepMind)	n/a	DeepMind (2024)
Junior office / admin worker	Document automation, support chatbots	30–50% of tasks	Slack WFI (2025)

Cashiers and administrative workers face the highest level of risk — self-service technology and OCR systems are already widely available and deployed by thousands of companies worldwide. In their case, automation is not a question of “if” but “when”.

Case Studies from Poland

PKO Bank Polski – voicebots in customer service

PKO Bank Polski, serving over 11 million customers, deployed a fleet of 17 voice bots that have collectively conducted more than 70 million conversations and saved over 1 million hours of consultant work. The bots handle enquiries 24/7, with no queues and no holidays. This is one of the largest customer service automation projects in Polish banking.

Tech-Metal 24 – invoice processing automation

Tech-Metal 24, operating in the metal industry, implemented a system combining OCR, RPA and a large language model (LLM) for automatic invoice processing. The results were measurable and swift: invoice processing time was reduced from 21 to 7 days (a 67% reduction), 650 working hours were saved annually, and yearly savings amounted to 123,000 PLN. The project’s ROI reached 85% — the company recouped its investment within a single year.

What to Do Now? A Checklist

Automation does not have to be a threat — provided the organisation responds proactively. Here are eight steps worth taking without delay:

1. Audit task vulnerability — identify which activities in your company are repetitive, rule-based and could be taken over by AI.
2. Plan for reskilling — develop career paths for employees at risk of automation: training programmes, new roles, internal promotions.
3. Deploy AI gradually — start with pilot projects in one department or process before scaling up.
4. Monitor and evaluate — measure the effects of automation: time, costs, quality, and customer and employee satisfaction.
5. Support employees — communicate changes well in advance and provide psychological and legal support where headcount reductions occur.
6. Build organisational capabilities — invest in IT and data analytics departments that will manage the deployed AI tools.
7. Comply with regulations (AI Act, GDPR) — ensure that AI deployments conform to European regulations on transparency and personal data protection.
8. Keep learning continuously — the AI market changes rapidly; follow new tools, benchmarks and industry reports to avoid being left behind.

AI Tools That Are Replacing Human Work

Many tools are already available on the market today that can take over part of the duties of office workers, content creators and customer service departments. Below is an overview of the most important ones.

Tool	Application	Cost / Pricing
ChatGPT / GPT-4 (OpenAI)	Text generation, chatbots, data analysis	Business plan: from approx. $20/mo; free basic tier
Google Gemini	Conversational assistant, information search	Free (Google account required)
UiPath / Blue Prism	Business process automation (RPA)	Licences from several hundred to thousands USD/yr
Midjourney / DALL·E	Marketing graphics generation	From approx. $10/mo
Synthesia	AI avatar video creation	Plans from $29/mo

Many of these tools offer free trial versions or freemium plans, allowing organisations to evaluate their usefulness before committing to a paid licence. The key to success is not the deployment itself but measuring ROI — how many working hours the tool saves per week and how quickly it pays back the cost of the subscription.

Ethics, Regulation and Security

From 2026, the AI Act applies across the entire European Union. It regulates the use of artificial intelligence systems in high-risk areas — including recruitment, credit scoring and medical diagnostics. Companies must ensure algorithmic transparency and the ability to appeal decisions made by AI.

The right to an explanation is one of the key elements of the new legislation. Employees have the right to know why an algorithm made a decision affecting their employment — for example during automated CV screening. Organisations that fail to provide this transparency risk financial penalties.

Data security remains a priority. Deploying AI tools involves processing large volumes of personal data belonging to customers and employees. GDPR requires companies to minimise data collection, encrypt it and report any breach. Before choosing an AI provider, check where they store data and whether they meet European standards.

Automation brings both risks and benefits. On one hand it eliminates certain jobs; on the other it creates new positions requiring higher competencies: AI engineers, data analysts, algorithmic ethics specialists. Organisations that treat automation as an opportunity to increase the added value delivered by employees — rather than simply a cost-cutting mechanism — will gain a long-term competitive advantage.

It is also worth remembering that some areas remain beyond the reach of effective automation — at least in the near term. Complex medical cases requiring empathy and holistic patient assessment, legal proceedings based on precedent, and social work requiring the building of trust relationships — these are fields where humans remain irreplaceable.

Summary and Conclusions

AI-driven automation is not a forecast — it is a reality already reshaping the Polish labour market. Cashiers, telemarketers, office workers and warehouse staff are in the front line of risk, but the pressure of automation will sooner or later affect almost every industry. The key question is no longer “will AI change my job” but “how quickly, and what will I do to prepare”.

For workers, the recommendation is clear: invest in skills that AI cannot easily replicate — critical thinking, creativity, emotional intelligence, project management and the technical competencies needed to work with AI tools rather than be replaced by them. Online courses, employer-funded reskilling programmes and industry certifications are the best insurance policy for the labour market of the future.

For companies and organisations the path is similar: proactive process analysis, gradual AI tool deployment, transparent communication with employees and full regulatory compliance. Automation can be a growth engine — provided it is implemented responsibly.

Ready for the automation era? Build your internal action plan today: audit your tasks, select AI tools to pilot, and schedule training for your team.