The new European Regulation is expected to come into force in 2024 and will help reduce fragmentation and create an environment of trust around AI. However, it will not fill a “regulatory gap” because measures such as the Regulation on the protection of personal data already affect the topic. In concrete terms, it is up to individual companies to create a risk management framework.
Artificial Intelligence (AI), its impact on society, and the close relationship and complementarity of regulatory and ethical aspects constitute a complex and still fragmented scenario, increasingly focused on the need to regulate AI, taking into consideration the various elements that characterize it and the challenges it poses.
The AI Act And The Current Regulatory Environment
The ethics of Artificial Intelligence concerns its economic, social and cultural implications, i.e. the repercussions on the health or safety of people and their fundamental rights. The topic is now the subject of great attention, so much so that in June, the European Parliament approved the Regulation called the AI Act, intended to regulate its use in the European Union. Final approval should be received by the end of the year for entry into force in 2024.
As stated in the text, the objective is to ensure the smooth functioning of the internal market for artificial intelligence systems… contribute to the Union’s aim of being a world leader in the development of safe, reliable and ethical AI, and ensure the protection of moral principles.
The proposed AI Act is developing in a legal context that is not characterized by a “regulatory vacuum”. The existence of European and national regulations (e.g. the Personal Data Protection Regulation, GDPR ) impacts several aspects of artificial intelligence.
The role of the AI Act, therefore, should be to constitute a tool aimed at avoiding excessive fragmentation and creating a context of trust around AI itself. In this scenario, the importance of soft law also stands out, which must be a support tool for legislation. An example is the code of conduct being discussed in the G7 or the role of the ethical guidelines published a few years ago by the independent high-level group of experts on artificial intelligence, established by the European Commission in June 2018, with the aim to promote trustworthy AI.
During the meeting it emerged that many obligations present in the AI Act proposal are apparently new but, in reality, already belong to the modus operandi of the jurist specializing in data protection. An example among all is the impact assessment on fundamental human rights, but also the risk-based approach according to which the greater the risk presented by the artificial intelligence system, the greater the attention paid by the legislator.
Lastly, in the current regulatory context, pending the approval of the AI Act, the need to coordinate national and local initiatives along the lines of Brussels emerges more and more clearly to avoid further fragmentation.
Also Read: Horizon 6G: The Role Of Artificial Intelligence
Ethics Of Artificial Intelligence: The Different Approaches At An International Level
Regulatory fragmentation is emerging at an international level. This is already made evident by the different approach taken by the European Union compared to that of the United States and China, which, although not too distant, are inspired by different modus operandi.
Furthermore, in the primary attention that legislators are paying to the protection of fundamental principles, cultural differences emerge between the various nations of the world, which, consequently, find themselves protecting some areas rather than others. In a heterogeneous context, the European Union pursues the aim of presenting itself as a promoter of values through the issuing of a binding rule, unlike, for example, the United States, which has chosen an approach based more on voluntary commitments of the actors involved.
The task of the European Union, therefore, is to not give up on principles even in the face of a more pragmatic approach such as that of the United States. To reconcile this discrepancy, dialogue and cooperation between the actors involved at an international level would be necessary, also taking into account the circumstance of the global social impact that AI systems have.
The Role Of Ethics As Soft Law
The theme of ethics was at the center of the meeting. The AI Act proposal has placed a lot of attention on the topic, starting from the assumption that technology must be developed and used for the good of the community. What emerges is a close correlation between ethics and law, which, however, does not translate into an overlap but is inserted into an already existing cultural context of values and principles.
Ethics can be understood as a soft law tool that allows the company to account for its decisions and to be responsible for any damage caused due to the lack (or poor) ability to anticipate and prevent (or mitigate) a foreseeable event that may harm the principles and values of the European Union. Ethics, therefore, embodies the concept of accountability by acting as a criterion of “correctness”. The importance of ethics is also attributable to the social, technological and legal evolution of recent times.
Following technological progress, we have witnessed a crisis of the principle of legality and a loss of centrality of law in which the latter no longer claims to organize and regulate the world systematically. In this challenging context, the use of soft law is justified by the motivated need to fill regulatory gaps. The reasons for the use of this legislative technique (soft law) can be traced back to the current need to adopt regulatory instruments with greater flexibility and, as such, capable of being adapted more quickly to the continuous evolution to which some areas are subject.
In concrete terms, it will therefore be up to individual companies to create an AI system risk management framework that protects the principles and values of the European Union to identify, estimate and evaluate the known and reasonably foreseeable risks that the technology, placed in a specific context organizational, may affect the health or safety of natural persons and their fundamental rights.
The Impacts Of Artificial Intelligence And Enforcement
The common feature of the use of AI systems is the possible impact on human rights. A fundamental problem concerns the transparency and justification of the training capacity of the algorithm, often fed by big data also obtained through the use of web scraping, which does not fully coincide with the principles established by the Regulation on the protection of personal data and, in particular, with respect for the rights of interested parties.
It is, therefore, necessary to create a system that aims to prevent risks (for example, the surveillance of AI systems or transparency) and not just contain them. To this end, it is essential to spread a sense of trust and encourage the development of skills, also among users of AI systems.
The current scenario is moving towards a regulation that will have local authorities as its enforcement landing point, and this could create problems in the application phase of the rules with a view to fragmentation already at a European level. The problem could be even more complex on a global scale if it continues in the absence of dialogue.
Conclusions
AI is a topic as complex as it is crucial, which is already having a considerable impact on everyday life. While waiting for a definitive solution, the current regulatory system and ethics (understood as a soft law tool) can represent a necessary tool to address complex problems that require pragmatic responses and articulated actions.
Also Read: Artificial Intelligence In Marketing, What CMOs Need To Know
