On March 24th, 2022, Prof. Thomas Hildebrandt and Prof. Henrik Olsen gave a presentation on AI and Legal Tech: Ethical Perspectives during the 20th session of the TUM IEAI Speaker Series.
In the lecture, the speakers presented on how new AI-powered technologies have changed the way that legal information can be analyzed, through network analysis, machine learning (ML) systems and the transfer of the law into systems through code. For instance, network analysis can help to cluster legal cases and display how they are interconnected. Moreover, ML-based natural language processing (NLP) systems could significantly help to increase the accuracy for predicting the likely outcomes of human-rights cases, helping to better use scare human resources for non-profits and government.
Artificial intelligence can also significantly help in having a more fair interpretation of law, as opposed to common manual methods that enable the tendency of interpreting the law in different ways. Often, automated processes’ interpretation of the law can lead to discrimination of vulnerable minorities. Therefore, there is a need to move from the automation of standard routines to models that navigate in complex and changing environments.
Rule-based AI can enable the handling of the complexity of the law by creating transparent digital law that can be used to provide treatment that is more equal. In practice, such models are helpful in annotating one model of the law that can be applied in many domains, such as dynamic guidelines for the users or compliant-by-design process support for caseworkers and employees. Such tools can enable a sustainable development of the legal tech field, by supporting the decision-making process with an overview of the necessary steps of compliance with increasingly complex legislation.
The session ended with a discussion of the recent EU AI Act and its implications. An exciting discussion followed on turning policy into practice and how to promote AI in the legal sector while minimizing ethical risks. We thank Prof. Thomas Hildebrandt and Prof. Henrik Olsen for their time and extremely interesting insights.
