There was a huge outcry in the US in the 1970s: it was claimed that cash machines would replace all bank tellers in the near future. ‘Machines will beat humans,’ was the refrain even back then. But things turned out quite differently, as Andreas Hopfener, Public Sector Lead for Insights and Data at the consultancy firm Capgemini, explained in his keynote speech ‘10 Future Theses on AI in the Public Sector and Their Impact on the Labour Market’. Forty years later, in 2010, there were 600,000 bank advisers in the US – twice as many as before. It was just that their job roles had changed.

“The question is: if intelligence becomes cheap, what will then become scarce – and therefore valuable? That’s where interesting opportunities arise,” says Hopfener. Thanks to AI, fewer staff are needed per task, but not overall. But who will be needed, and where? The consultant has put forward ten future scenarios for the jobs of tomorrow:

1. Demographics are making AI a necessity for the public sector, as there is a shortage of skilled workers.
2. Intelligence is becoming a commodity, i.e. a mass-market product. And cheap intelligence generates more demand, not less.
3. Software is becoming cheap, but integrating it remains expensive. This increases the need for specialists with domain-specific knowledge and expertise in areas such as data management, architecture and risk management for AI systems. In the public sector, existing backlogs are creating additional demand.
4. Agents are replacing the traditional software interface. Now, people set a goal for the agent rather than simply operating functions.
5. New roles are emerging. Wherever AI agents are deployed, someone must develop, test and take responsibility for them.
6. Trust-by-design is becoming a purchasing criterion and creating a labour market of its own. When agents process sensitive data, governance requirements increase. This calls for experts.
7. AI is becoming the infrastructure underpinning the state’s performance and sovereignty, and thus a cross-cutting issue for public administration, public services and the state’s capacity to act. The shortage of skilled workers is the greatest obstacle to digital sovereignty.
8. AI only has an impact when the organisation changes. Integrating it into day-to-day operations remains a major challenge.
9. AI initially takes over the tasks that newcomers have traditionally used to learn the ropes. New learning pathways must therefore be developed.
10. Use cases take precedence over models: value is created within the specialist domain. This requires experts who understand the problems at hand.

The expert’s overarching argument is therefore that AI is transforming the public sector. Skills shortages and shifting career priorities are opening up new job opportunities.