Challenges of Artificial Intelligence, the search for balance
In Brazil, the significant amount of legal data available in public access creates fertile ground for the emergence of innovations.
The first thematic axis of publications related to Artificial Intelligence (AI) and its impacts in different social sectors ends, from the perspective of the connection between law and technology. Reflections and provocations were shared regarding the developments of these innovations in the Judiciary, economic sectors, more appropriate solution of conflicts, diversification of players in the legal services, together with the national and international regulatory discussion on conduct parameters aimed at the parties involved with the development and application of AI systems.
The data economy witnesses the consolidation of this new asset with economic value, driven by the growth in the volume of data and the possibility of massive collection from varied and unstructured sources. Thus, the importance of a definition is emphasized, for normative and regulatory purposes, that focuses on the set of rules and instructions provided to those systems in programming language (algorithms) and on the set of information used in the design, training and availability of subsequent products and services (data).
In Brazil, the significant amount of legal data available in public access creates fertile ground for the emergence of innovations. On the one hand, there is a multiplication of applications that intend, as the case may be, to digitize services and activities, automation, classification and document review, compliance, education, e-discovery, management, jurimetry, legal research, Online Dispute Resolution, virtual assistants, predictive analysis. On the other hand, the stakeholders expect results, such as, for example, saving time and resources, risk management, improving routines and procedures, preventing conflicts, reducing repetitive tasks.
This is the breadth of systems, products and services that demand caution when proposing rules and regulations. National and international recommendations are divided into five main topics: (1) Guarantees for human persons – autonomy, well-being, equity, inclusion, sustainability; (2) Technical requirements – approach based on risk, reliability, representative and diverse data set, explainability, human intervention, possibility of auditing by independent third parties, accountability, protection, accountability, robustness, security, transparency; (3) Legal compliance – fundamental and human rights, rule of law, non-discrimination, privacy, protection of personal data, democratic values; (4) Performance of state actors – training of people, international cooperation, data governance structure, facilitating the participation of the private sector, investment in R&D; (5) Mechanisms for checking compliance.
The agenda of the regulatory debate is therefore oriented towards the search for a balance between innovation, economic competitiveness, broad benefits to society, protection of fundamental rights and democratic values. This can be achieved through flexible rules and regulations in a manner consistent with rapid technological developments, regulatory impacts, potential risks from AI systems, legal certainty and the appropriate liability regime. It is therefore expected that the Brazilian AI initiative will be organized based on the dialogue, participation and negotiation of all interested parties.
By: Wilson Sales Belchior