We don’t want to say that artificial intelligence is becoming exceedingly popular and that it is occupying a very important and quite large place in our lives. You probably already know that, you probably even use artificial intelligence to draft briefs for your clients, to summarize huge texts when you just don’t have the time or to just come up with ideas for your next public speaking.
Therefore, you’ll probably want to be aware of the European Union’s steady progress in codifying the rules for using artificial intelligence. In the last two years the European Parliament and the Commision published for discussing the following proposals:
- Proposal for a Regulation of the European Parliament and of the Council laying down harmonized rules on artificial intelligence (Artificial Intelligence Act) and amending certain union legislative acts;
- Proposal for a Directive of the European Parliament and of the Council on adapting non-contractual civil liability rules to artificial intelligence (AI Liability Directive).
The text in this article aims to summarize arrangements proposed by the European Parliament and the Council and try to make clear both the constrains and the new horizons that will apply in the European Union in the near future and therefore for all Member States.
ARTIFICIAL INTELLIGENCE ACT (AIA)
The Artificial Intelligence Act will apply to both providers and users of AI systems. The biggest goal for the Act is to ensure there is transparency and accountability when using (professionally) or providing AI, an integral part of which is at the end human oversight.
Furthermore, compliance with data protection law in the EU shall be a must when providing or using AI. Risk assessment is also among the future regulations.
Keeping in mind those future principles, some AI systems shall be classified as high-risk and will be subject to stricter requirements. For instance, AI used in healthcare, transportation, and law enforcement shall fall into the high-risk category. Biometric identification in public spaces for law enforcement shall also be limited.
Despite putting the usage of AI in a frame, the Regulation also recognizes the importance of fostering innovation and helping smaller businesses compete in the AI sector. The AIA aims to achieve this with the following:
- SME-friendly requirements - the regulation is designed to have flexible and proportionate requirements;
- Support and Guidance - educational materials, training programs, access to expert advice, etc.;
- Innovation Incentives - stable and predictable environment for SMEs with clear regulatory framework;
- Market Opportunities - the AIA’s emphasis on AI safety and accountability shall create a marketplace for SMEs that focus on innovative, ethical and responsible AI solutions;
- Collaboration and Partnership, Access to Data, etc.
Thus, these measures not only aim to regulate the AI sector and impose sanctions on those who infringe those rules. Above all else, these measures are intended to help SMEs compete in the AI space and contribute to the overall growth and competitiveness in the industry.
AI LIABILITY DIRECTIVE
On a more serious topic, the AI Liability Directive aims to address the legal challenges and implications of AI systems that cause harm or damage. Whether you are using or being affected by AI, it is important to know who will be considered responsible for any harm caused by it according to the law. Of course, the Directive focuses primarily on non-contractual civil liability whether it is in cases of individuals’, property’s or environment’s harm.
The key provisions include the following:
- Strict liability - the operator of an AI system that causes harm shall be held liable without need to prove fault or negligence;
- Operator Responsibility - the operator (meaning the provider, the user, the authorized representative, the importer and the distributor) shall be considered the primarily responsible party. So, in cases of multiple operators, i.e. the developer, the owner and the user, a contractual agreement would be the best solution for distributing the responsibility;
- Cross-border liability - all victims shall be able to seek compensations regardless of the location of the AI operator;
- Compensations - the Directive intends to ensure that both material and immaterial damages be compensated;
- Duty to report incidents - all operators shall be obliged to report AI-caused incidents that result in harm or damages;
- Public Access to Information - the Directive shall allow public access to information about AI system operators, making it easier for victims to find the responsible party.
- and many more.
It’s important to always remember that the Directive will have to be transposed in the national legislation of all Member States. This means that if someone wants to ask for money because they were hurt or harmed, they will have to go to a local office or organization in their own country. It might take a while for this to happen, but eventually, it will affect how we live our lives every day.
By drafting both the AIA and the AI Liability Directive, the European institutions aim to ensure balance between promoting innovation and making a path for SMEs, on the one hand, and protecting victims’ rights, on the other. If you find the topic interesting and want to find out more about how using AI professionally, don’t hesitate to contact us!
This article is up-to-date as of 01.11.2023 and is a brief summary of the future rules concerning AI on European level. It does not constitute legal advice or guidelines for conduct and therefore we do not advise undertakings to make commercial decisions based on it. For more information, you can contact us at office(at)recht.bg.