With the growing importance of Artificial Intelligence in management and the world of work, CEC European Managers highlights the need to rethink AI governance and leadership. To promote innovation capacity, social well-being and a positive impact of AI systems, better stakeholder engagement and involvement of social partner is needed. Against that background, CEC member CFE-CGC has conducted an EU project on promoting confidence in AI through dialogue. The final report is now available.
“To harvest the opportunities of AI for Europe, we need to improve our leadership and innovation capacity. Global competition requires finding more agile ways of innovating and regulating through better involvement of stakeholders like social partners in the whole value creation cycle” Maxime Legrand, President
The SECOIA-DEAL EU project, under the leadership of CEC member CFE-CGC in France, has been able to gather a wide range of European stakeholders to pioneer new models of value creation, governance, and regulation of Artificial Intelligence systems. The final report highlights concrete ways of promoting trust and transparency of AI systems across the value creation cycle, with a special focus on the role of the workplace and the role of managers in it.
In the context of upcoming risk-based EU regulation on AI, the project concludes that a complementary bottom-up approach can increase the innovation capacity in AI value creation cycles. Furthermore, involving social partners and communities of users in the development of AI systems can be a more agile way of governing AI ex-ante. This has wide implications for making AI’s impact a better fit to the EU’s economic, social and environmental priorities.
“The European Union wants to have a comprehensive legal framework to limit the drifts of Artificial Intelligence without constraining innovation too much. It is clear that this road is long and winding, and the arrival of ChatGPT and so-called generative AIs have disrupted an already constrained schedule. It was therefore also necessary for the social partners to anticipate the impacts of AI on jobs and the world of work, by imagining the social dialogue that will complement this future regulation. This is what we have done with the European project SecoIA Deal, which produces tools and concrete proposals in order to be able to respond in companies.” Nicolas Blanc, expert in Artificial Intelligence from CFE-CGC and CEC emphasises.
The final report recommends concrete measures on improving AI governance and dialogue, including:
- AI Tools record: AI projects of organisations shall be recorded with a description and evaluation of the projects. This record will allow for early involvement of all stakeholders in the organisation to strengthen accountability through collective decision making and monitoring.
- Review clause in the framework of the cycle of use
of AI-based tools at work: In order to ensure relevance and guarantee confidence in the tool and in the purpose of its use, it seemed useful to imagine a “permanent”, “long-term” social dialogue on AI-based tools used within the company. This dialogue would be based, among other things, on the support of a review clause.
- Corporate AI Ethics Committee : An Ethics Committee can ensure to maintain the trust relationship between the provider and the user, which is based on the provider’s accountability and the ability to have AI referents in the company. The creation of an Ethics Committee involving employee representatives, in conjunction with the proposed creation of an AI Ethics Ofcer and record-keeping, is likely to strengthen users’ obligatons and the consideraton of the evolving nature of AI systems.
- Stakeholder Value Radar : The Stakeholder Value Radar applicaton will verify that the value created by the application is equitable, and that there are no winners and losers among the stakeholders. In addition, the tool will facilitate the awareness of all stakeholders of the positive and negative aspects of the application of the IA, thus creating a symmetry of information.
Read the report