Following various policy papers and guidelines published by high-level working groups on Artificial Intelligence ("AI"), on the 21st of April 2021 the European Commission issued a set of proposals for the introduction of harmonised rules on artificial intelligence ("Artificial Intelligence Act") and for the amendment of current EU legislation to bring this in line with the new rules. These proposals come in the wake of several concerns raised over the last years regarding the risks and/or negative impact that AI can have on rights and freedoms of individuals and on society at large1.
Whilst the European Community ("Community") recognises the 'wide array of economic and societal benefits across the entire spectrum of industries and social activities' that AI brings with it, the introduction of this proposed regulatory framework aims to achieve a level legal certainty within the AI industry, to ensure that AI systems are safe and respect European Union ("EU") laws and values, to enhance governance and enforcement of existing laws, fundamental rights and applicable safety requirements, to facilitate the development of a single market for lawful, safe and trustworthy AI applications and to prevent market fragmentation2.
The Community is seeking to achieve these objectives by introducing balanced and proportionate rules, providing for minimum requirements to address the risks and problems linked to AI usage and without stifling technological development.
Following a public consultation launched earlier last year, during which 1251 submissions were made, the following proposals were put together by the European Commission:
- 'single future-proof' definition of AI,
- harmonised rules for development, placement on the market and use of AI,
- restriction on legal intervention to be limited to justifiable situations,
- prohibitions on certain harmful AI practices,
- specific restrictions and safeguards for biometric identification for purposes of law enforcement,
- risk methodology to define "high risk" AI systems particularly those that pose health and safety risks or risks to fundamental rights,
- compliance to mandatory requirements to ensure trustworthy AI's,
- conformity to procedures prior to an AI product being placed on the market.
The proposed regulations also introduce a set of obligations on providers and users alike, to comply with existing legislation throughout the AI's lifecycle.
Similar to the approach adopted by the Community in regulating other 'new technologies' launched in the last decade, such as crypto currencies, the proposed rules will be enforced through a governance system at Member State ("MS") level, building on already existing structures, whilst introducing a cooperation mechanism at Union level through the establishment of a European Artificial Intelligence Board ("EAIB") and regulatory sandboxes.
The proposed rules complement and require full consistency with existing Union legislation including, but not limited to, the EU Charter of Fundamental Rights, competition law, data protection law and consumer protection law. The proposals also cater for rules pertaining to high-risk AI systems which will also be integrated into existing sectoral safety legislation to ensure consistency and avoid and/or minimise any duplications and burdens.
The proposals suggest that designated supervisory authorities will have to be appointed in order to take charge of implementing these legislative requirements in a manner which ensures that the necessary technical expertise and resources are available. The Community is also proposing that existing regulatory authorities are entrusted with the supervision of regulated entities, such as credit institutions and financial services, to supervise and ensure that a coherent enforcement of the proposed obligations is observed. On a local level, the Malta Financial Services Authority ("MFSA") recently issued a circular3 to gather feedback on how these proposed new rules might affect stakeholders in their operations and compliance requirements, and whether the implementation of AI systems within their businesses is foreseeable.
Although the proposals are likely to provoke a wide range of reactions within the various economic sectors of the Community which will be impacted by the regulation of AI, it is our view that any legislative developments and amendments which aim to properly regulate the use and exploitation of new technological systems without stifling innovation and fair competition, are a step in the right direction and will certainly provide the necessary foundations for AI systems to be 'a force for good in society and increase human well-being'4.
1. European Commission, 'Proposal for a Regulation of the European Parliament and of the Council, Laying down harmonised rules on artificial intelligence (Artificial Intelligence Act) and amending certain Union Legislative Acts'.
2. ibid p. 2-3
Originally Published 25 May 2021
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.