Forthcoming European Commission Communication on Artificial Intelligence: ETUC CALLS FOR AN AMBITIOUS STRATEGY RESPECTFUL OF WORKERS
Ahead of the upcoming European Commission’s 2020 white paper on artificial intelligence, the European Trade Union Confederation reiterates the need for an ambitious European AI strategy to maintain and reinforce workers’ protection and involvement, and empower them to create sustainable use of AI tools.
AI systems are data-driven technologies. Access to, and the ownership of, data are the core of AI technologies. Data has created a new business model for companies. However, the boundaries between private and non-private data are thin.
Data is sensitive. AI innovations are not per se good and do not per se deliver positive outcomes for society. Access to and processing of data needs regulation for legal certainty and predictability, security and safety, and protection for all. Ethical principles are key. They should form a robust and reliable basis for business, workers and society. Ethical principles should be legally binding. Only under this condition will they provide a level playing field and fair competition. However, one AI regulation cannot fit all situations: consumer protection and worker protection need a differentiated approach.
An ambitious European AI regulatory framework should address the specificity of the workplace. Humans must be in command. Any AI technology should enable humans remain in control. Workers must be able to opt out from the human-machine. The regulations must specifically address workers’ data protection and privacy and go beyond GDPR.
Digitals skill are crucial. Workers need to be empowered and critically aware of what AI technology at work brings. They need to become “AI literate”. GDPR is a powerful tool that trade unions can use to exercise the “right to explanation”. Worker representatives should have a major role in ensuring this right at the workplace.
AI regulations should prevent disproportionate and illegal surveillance at work. They should prohibit discriminatory treatment on the basis of biased algorithms. Evidence shows that workers’ consent is usually forced and cannot equal an informed consent due to the inbalance of power in the employment relation.
Rules on liability will become crucial. The reversal of the burden of proof will be required to provide a balanced approach to risks. The EC will certainly need to revisit the traditional liability regime given the complexity of the technologies at stake.
An ambitious European AI regulatory framework should grant workers the right to data governance at company level. The ETUC calls for strengthening workers’ involvement in the design, deployment and use of AI technology. ETUC calls for reinforcement of the right to timely and useful information and consultation when new technologies are planned and introduced at the workplace.
AI regulation should identify and mitigate work-related impacts of AI technologies, as much as economic and environmental impacts. Public and private investment will be needed. Public investment is required in those countries and regions lagging behind on the road to digitalisation.
AI needs a legally and empowering European framework based on human rights, public interest at the service of society, for the social and environmental wellbeing and common good. AI technologies will only deliver a fit for purpose innovation, if they comply with the Treaty based precautionary principle.
The ETUC strongly believes it is dangerous to allow AI technology to be allowed onto the market without being properly tested. Regulation from the financial services industry allowing untested products to be released onto the market is totally inappropriate for AI.