In the European Union (EU) and Europe in general, there is currently no legally binding definition of artificial intelligence (AI) nor a common EU framework for the regulation of AI. The EU Independent High Level Expert Group on Artificial Intelligence (AI HLEG) defines AI systems as “software (and possibly also hardware) systems designed by humans that, given a complex goal, act in the physical or digital dimension by perceiving their environment through data acquisition, interpreting the collected structured or unstructured data, reasoning on the knowledge, or processing the information, derived from this data and deciding the best action(s) to take to achieve the given goal.”Footnote ¹ The European Commission in its white paper emphasized that a definition of AI “will need to be sufficiently flexible to accommodate technical progress while being precise enough to provide the necessary legal certainty.”Footnote ² In particular, the terms “data” and “algorithm” need to be defined. With the exception of autonomous vehicles, there are not many AI-specific binding rules yet. A common EU legal framework, which has been proposed in the European Commission's white paper, will build trust and provide legal certainty for consumers and businesses.

The Council of Europe and the EU have produced many documents that deal with the impact of AI on human rights. These documents have in common that they embrace a human-centric approach to AI, in particular human oversight and agency. AI systems should enable human rights by empowering human beings, allowing them to make informed decisions and fostering their fundamental rights.Footnote ³ This principle is also at the core of Article 22 of the General Data Protection Regulation (GDPR) of the EU.Footnote ⁴ Article 22 on automated individual decision making states that “[t]he data subject shall have the right not to be subject to a decision based solely on automated processing, including profiling, which produces legal effects concerning him or her or similarly significantly affects him or her.”Footnote ⁵ “Solely” means that the decision is totally automated and there is no human review. “Legal or similarly significant effects” means that the decision affects a person's legal status or rights or it is something that has an equivalent impact on an individual's circumstances, behavior, choices, or leads to exclusion/discrimination of the individual. Examples are social benefits, voting, online credit application, or university admissions.Footnote ⁶

In April 2019, the AI HLEG released its final “Ethics Guidelines for Trustworthy Artificial Intelligence.”Footnote ⁷ In these ethics guidelines, the expert group proposed seven principles to design trustworthy AI based on fundamental rights. These are: (1) human agency and oversight; (2) robustness and safety; (3) privacy and data governance; (4) transparency; (5) diversity, non-discrimination, and fairness; (6) societal and environmental well-being; and (7) accountability. In June 2019, the AI HLEG published further policy recommendations for EU and national policymakers and also launched a pilot phase for testing of the ethics guidelines by organizations, which was concluded in June 2020.Footnote ⁸

The European Commission in its White Paper on Artificial Intelligence released in February 2020 proposed a risk-based regulation for AI.Footnote ⁹ High-risk applications are generally defined by the sector (such as health care, transport, energy, and parts of the public sector) and the intended use, meaning the impact on the affected parties. However, some applications are always considered high-risk for fundamental rights violations and will therefore always be subject to the proposed rules. These areas are employment (recruitment), biometric identification (facial recognition), and other intrusive surveillance technologies. Developers of non-high-risk AI systems may voluntarily comply with the rules and receive a “quality label.”

With regard to AI and its impact on human rights, a highly anticipated decision from the District Court of The Hague, Netherlands decided on February 5, 2020 should be mentioned.Footnote ¹⁰ The case concerned the growing use by governments of AI and risk modeling in administering welfare benefits and other services. The Court held that the System Risk Indication (SyRI) algorithm system, a legal instrument that the Dutch government uses to detect fraud in areas such as benefits, allowances, and taxes, violated Article 8 of the European Convention on Human Rights (right to respect for private and family life). It stated that the damage caused by welfare fraud justified the conclusion of the legislator that there was a compelling social need to take measures in the interest of the economic well-being of the Netherlands. However, the operation and the associated procedures and safeguards of the SyRI were insufficient and the SyRI legislation did not strike a fair balance. Furthermore, the principle of transparency was not observed, because there was no insight into the risk indicators and the operation of the risk model, according to the Court.