Since 2019, the French government has provided Pôle emploi with a substantial budget to "install artificial intelligence as a means of accelerating the return to sustainable employment". Within the framework of the transformation of public action, this fund called "Employment Intelligence" aims to improve the efficiency of Pôle Emploi and to make agents' missions simpler. As part of this project, in July 2020, the International Service of Pôle emploi (Directorate Europe and International Relations) launched a consultation of European public employment services to obtain an overview of initiatives on artificial intelligence.
This consultation served a twofold purpose:
•To have an overall view of the initiatives carried out by the European public employment services in the field of artificial intelligence: impact of AI on both economy and employment, adopted methodology to transform itself through AI, AI solutions implemented, etc.
•To gather the different partners' preferences to define a possible cooperation framework.
The consultation consisted of a questionnaire addressed to 30 PES European partners.
A first analysis of this survey's feedback reveals that most EU Member States have developed strategies on the use of AI in PES. Work focuses mainly on the evolution of the labour market, the development of new skills and the implementation of predictive tools. The Artificial Intelligence vision for PES is particularly centred on decision support.
The emerging conclusions of this broad survey mainly relate to 5 areas:
1.Ethics emerges as a key issue for the different PES involved:
Most PES have conducted studies on ethics, however advanced and experienced they are on AI topics: everyone agrees that the definition of an ethical framework for AI constitutes an essential prerequisite towards AI transformation.
The PES agree that AI tools must remain subordinated to human beings and must ensure better support for jobseekers.
2.HR impact & change implementation: Several trends can be identified in terms of human resources:
Regarding trades, we are witnessing a gradual shift towards more specialised technical roles: AI platform engineers, Machine Learning (ML) engineers, AI-Backend engineers and AI/ML architects, etc., as well as maintaining a wide range of skills: data engineers, data scientists, machine learning experts, project managers,
We are also observing an evolution in the agents' / advisers' (or even to a greater extent) training to encompass the value of the data and the related actions.
Change management means that the whole organisation must become familiar with AI and that users must be involved in the design of solutions. Soft skills such as empathy, flexibility, creativity and strategy development are central to the development of artificial intelligence.
3.External relations & governance: Other public agencies and administrations remain the privileged partners of the PES:
The first partners of the PES in AI development consist of other government agencies and public administrations, either at national or regional level according to the country. This collaboration particularly relates to the issue of data sharing and should continue to grow. Several PES also report partners from the Research sector, universities and institutes. Companies mainly contribute by providing expertise in the development of AI implementation (e.g. language model development). No formal strategy seems to have been developed yet regarding the GAFAMs, which rather represent a source of inspiration. Some mention the intention to collaborate with them as well as with other private companies to share data, which implies, at this stage, another cultural step.
4.Data management and input appear as major concerns of the different PES.
For the individual PES, both the volume and quality of available data are fundamental to the development of relevant solutions. These are identified by several PES as the first key success factor for a successful AI transformation process. However, concrete data-sharing cooperation with various partners is still limited or only relates to non-personal data (e.g. labour market data).
Compliance with the General Data Protection Regulation (GDPR) is considered by everyone as essential but also as a source of complexity in the development of AI use, and particularly in data sharing.
Data structuring work is becoming more and more important with an increased need for professional expertise.
5.AI solutions & results: AI's benefits are considered promising, although little quantitative measurement has been made so far.
The PES engaged in research in AI, are globally developing AI solutions aimed at improving knowledge of the labour market, carrying out diagnoses and predicting at-risk groups, as well as matching supply and demand.
To date, the results still seem to be poorly measured quantitatively, and not in a comprehensive manner.
The French PES, Pôle Emploi, will soon initiate bilateral meetings with other European PES particularly committed or involved in this approach or on specific themes, such as ethics, ontology for the labour market, data management, ...
It plans to currently extend this benchmark to other PES on other continents (Asia-Oceania, Africa, America,) and initiate exchanges with international organisations (ILO, WAPES, UNESCO, World Economic Forum, OECD,...).
Your contact person :
Chargé de relations et actions internationales
Direction Europe et Relations internationales
Direction Générale de Pôle emploi