Pro-worker AI: levers for harnessing the potential and mitigating the risks of AI in connection with employment and labour market policies

Download — Становище на ЕИСК: Pro-worker AI: levers for harnessing the potential and mitigating the risks of AI in connection with employment and labour market policies

Key points

The EESC:

  • considers that social dialogue and worker involvement play a crucial role in preserving workers’ fundamental rights and promoting ‘trustworthy’ AI in the world of work. One of the levers for minimising the risks and harmful impacts of AI systems is stronger involvement of workers and their representatives;
  • insists that legislative initiatives, or any other initiatives adapting existing law, should address the gaps in the protection of workers’ rights at work and ensure that humans remain in control in all human-machine interactions;
  • considers that the EU’s existing legal provisions relevant to the use of AI in the workplace should be backed up by explicit guidance;
  • supports the swift implementation of Article 4 of the AI Act to enable employees to manage AI tools;
  • considers that public authorities must implement skills development initiatives for workers and citizens to ensure that artificial intelligence systems enhance, rather than replace, humans;
  • issues a strong call for enforced social dialogue on the deployment of AI systems on the basis of an ad hoc EU legal instrument that includes provisions to achieve the following more effectively:
  • enable the enforcement of Article 88 of the General Data Protection Regulation (GDPR)
  • broaden the scope of the provisions contained in the Platform Work Directive (PWD), addressing the challenges that algorithmic management systems pose to all workers;
  • strengthen the rules applicable under Directive 2002/14/EC when high-risk AI systems are introduced, and give explicit guidance on the provisions of EU Directive 89/391 on Safety and Health at Work;
  • extend the communication of Data Protection Impact Assessments (DPIA) to workers’ representatives, as provided for under the PWD;
  • provide ex-ante Fundamental Rights Impact Assessments (FRIA), to be carried out by providers before high-risk systems are deployed; and
  • establish clear guidelines on how sandboxes and real-world testing conditions can be used.
  • Due to divergent views, members of the EESC Employers' Group tabled a counter-opinion. The counter-opinion was rejected by the EESC (112 votes in favour/136 against/11 abstentions) but since it received more than a quarter of the votes cast, in line with the EESC Rules of Procedure, it is appended to the opinion.

Downloads

  • soc/803_Record of proceedings