AI is changing the world of work. It will become increasingly important in public administration as well. This raises new questions, some of which will need completely new answers. This is why it is all the more important to open the floor to debate about the possibilities and opportunities – and also the challenges – associated with the use of artificial intelligence. For those reasons, the Policy Lab Digital, Work & Society is building a network on AI in labour and social administration, and is launching a series of expert discussions on the topic. Nine workshops, known as AI Labs, will be held in spring 2022, where experts from government agencies involved in labour and social administration will share their thoughts and work together to develop self-regulatory guidelines for their work in practice. The participants will include, among others, representatives from the statutory pension and accident insurance bodies or the Federal Employment Agency.

The expert discussions will focus on questions and considerations around responsible use of intelligent systems. What will AI be able to do for administration in the future? What are the benefits to staff members and members of the public? How can data protection, transparency, safety and security be ensured? And how can changes be designed to keep social considerations and people at the heart of administrative operations? Does the use of intelligent systems need to be limited and controlled? The expert discussions will aim to develop ideas for how artificial intelligence can be used as a practical tool in a sustainable way to achieve better and more efficient government administration. The experts will also share their thoughts on the challenges faced and will discuss ways to tackle them. Ultimately, the aim is to come to a shared consensus on the responsible use of AI, which can be used to guide and direct administrative action.

AI holds huge potential for improving quality, increasing efficiency and saving time for members of the public

It was clear in AI Lab #1, which was held on 4 May 2021, that the participants saw huge potential in the use of artificial intelligence. To give several examples: in the future, machines will be able to take over routine tasks, reducing the workload of staff members and giving them more freedom and time to devote to advising members of the public. Intelligent systems are able to analyse large amounts of data and detect patterns, retrieve documents and text, or point out potential problems – significantly reducing processing times, which means that wait times for members of the public will be shorter. The main emphasis is on how AI can assist in the often time-intensive preparation stage of the decision-making process. Having said this, no one in the network wants to outsource the entire decision-making process to AI systems.

Self-regulatory guidelines for practical use of AI in government administration

The series of expert discussions was opened by the Federal Minister of Labour and Social Affairs, Hubertus Heil, with AI Lab #1. As Heil asserted, ‘The use of artificial intelligence must be safe and secure. It must operate in a way that engenders trust and does not discriminate.’ For this reason, high standards must be set for its use in labour and social administration. The expert discussion series aims to develop these standards through rigorous debate and set forth specific self-regulatory guidelines for use in practice. The actual practical experience of the experts involved in the network will play an important role in the development of these standards and guidelines.