Instructional goals
This laboratory is the fourth stage of the overall didactic structure of Legal Informatics of the LUISS master's degree in Law, which is made up of five preparatory laboratories and a course spread over the first three years. These teaching classes constitute a unitary and progressive path according to a logical itinerary studied in order to implement the student's sensitivity for the mutual interaction between information technology and law. The intent is to train a jurist who can be ready to face the legal challenges of the digital dimension, increasingly pervasive and transversal in every professional sector, and of IT applications in the legal sector. To this end, the student will also have to acquire purely technical and IT knowledge to fully understand the technological phenomena of which he or she may be required to evaluate the legal implications and effects. This is a strategic goal that the Department of Law has set itself, as it is impossible to imagine the figure of a jurist today who is not fully familiar with digital tools and is unable to analyze the impact of the most disruptive technological applications on society, law, markets and institutions at a global level. The overwhelming innovation encouraged by national and European public policies requires versatile professional figures, capable of applying the traditional categories of law to unprecedented technological phenomena, or even of building new ones better able to regulate the present. To do this, it is required an understanding of the basic architecture of networks, as well as the languages of mathematics and the logic of algorithms, in order to be able to read them in the forms of law. Specifically, the Artificial Intelligence, Machine Learning and Law Laboratory serves as a direct complement to the immediately preceding laboratory, focusing on the legally relevant aspects concerning the implementation of AI. In particular, the course is aimed at: re-reading the general theory of artificial intelligence from a legal perspective, in the wake of the subject that still encompasses the teaching of legal informatics, i.e. philosophy of law; therefore to demonstrate the applicability of artificial intelligence algorithms to the various traditional functions of law; and finally to provide students with an overview of the main experiences of AI regulation at national, European and global level.
Prerequisites
Intelligenza artificiale.
Intended learning outcomes
Knowledge and comprehension: to achieve knowledge of some cutting-edge topics in the relevant field of study with the appropriate teaching support, understanding the difference between the logic of human intelligence and that of artificial intelligence in relation to the formation of a decision and the social and legal impact of AI applications, understanding the application of different types of algorithms for legally significant purposes and knowledge of the main regulatory instruments existing and proposed for the regulation of AI. Applied knowledge and comprehension: devising and supporting arguments relating on the one hand to the applications of AI in the legal sector and on the other to the major challenges of AI regulation. Making autonomous judgements: collecting and interpreting relevant information and data. Communication skills: communicating information, ideas, problems and solutions with the specific technical vocabulary of artificial intelligence. Ability to learn: having developed the skills necessary to undertake subsequent studies with a high degree of autonomy.
Course Contents
I. Theory of artificial intelligence and law
II. The different types of artificial intelligence and machine learning, fair division algorithms and the law
III. Law and regulation of artificial intelligence
Reference Books
The teaching material consists of the content of the lessons given by the lecturer, related handouts and other materials shared on Luiss Learn. Recommended readings:
Teaching Methods
Acquisition: lectures, podcasts and online quizzes Practice: guest speakers, case study and simulation Investigation: analyzing ideas and information in a range of materials and resources, using legal databases to collect and analyze data and comparing texts Collaboration: small group project, discussing others’ output and building joint output Discussion: seminars, group based class discussion, online forums and synchronous and asynchronous discussion Production: essays, reports, presentations and blogs
Assessment Method
The final grade, expressed out of 30, will derive from the evaluation of the following items for the respective percentage share:
20% attendance
10% active participation during classes
50% intermediate tests (no more than 3, with multiple-choice tests and/or open-ended questions)
20% final exam (oral)
N.B. The grade obtained at the outcome of the exam of this Laboratory will participate for the share of 1/7 in the final grade which will be attributed to the outcome of the exam of the Macchine intelligenti e diritto (MID1) course and which regularly falls within the curricular average grade of each student.
Thesis assignment criteria
Attendance at lessons, submission of project
work and successfully passing the oral
examination
Week 1
Module 1: Artificial Intelligence, Law, General Framework (Part One) 1. Introduction to artificial intelligence: functions, purposes and perspectives 2. Margaret Boden: types of AI 3. Focus: AI as a tool for knowledge. Scientific perspective and epistemological reflection 4. Basic notions of symbolic and sub-symbolic AI
Week 2
Module 1: Artificial Intelligence, Law, General Framework (Part Two) 1. A short history of AI paradigms: symbolic AI, cellular automata, genetic algorithms 2. Oliver Wendell Holmes and the idea of a "science-driven" jurisprudence 3. Focus: computation as a tool to understand and create law 4. Introduction to computational legal empiricism and computational social sciences
Week 3
Module 2: Agent-Based Modeling – Artificial Intelligence for the Study of Social Order (Part One) 1. Introduction to complexity theory: complex systems, emergence, non-linearity, phase transitions 2. Computational social science: data-driven and model-driven approaches 3. Simulation as a tool for understanding social complexity 4. Generative social science: the contribution of Joshua Epstein
Week 4
Module 2: Agent-Based Modeling – Artificial Intelligence for the Study of Social Order (Part Two) 1. Introduction to agent-based simulation: first examples with NetLogo 2. Thomas Schelling’s racial segregation model: structure, dynamics, interpretation 3. Social norms, norm evolution and agent-based models 4. Applied examples: Axelrod, sanctions, commons and genetic algorithms
Week 5
Module 3: Computational Crime Analysis: Artificial Intelligence and Computational Social Sciences in Criminal Justice (Part One) 1. Objectives and functions of criminal justice 2. AI applications in criminal justice: law enforcement and judicial support 3. Crime forecasting: hotspots, offenders, victims – the PredPol case 4. AI Driven Policing as an emerging research field 5. Predictive Risk Assessment and the COMPAS case 6. Basic introduction to machine learning mechanisms
Week 6
Module 3: Computational Crime Analysis: Artificial Intelligence and Computational Social Sciences in Criminal Justice (Part Two) 1. Critical issues: transparency, accountability, and data bias 2. Institutional positions: Council of Europe, European Union, AI Act 3. AI in criminal justice as a high-risk system 4. Beyond prediction: agent-based modeling for criminological theory (Charlotte Gerritsen) 5. The CrimeMiner project: network analysis, supervised classification, human-machine cooperative learning 6. Reflections on the use of AI for investigative, predictive and interpretive purposes
Week 7
Module 4: The Rule of AI – Artificial Intelligence and Technoregulation (Part One) 1. AI and the regulation of social life: technological power 2. Key perspectives: Cathy O’Neil and Benjamin Bratton 3. Introduction to the concept of technoregulation 4. Key concepts and authors: – Lex Informatica (Joel Reidenberg) – Code is Law (Lawrence Lessig) – Technological Management & Normative Environments (Roger Brownsword) – Code-Driven Normativity (Mireille Hildebrandt) – Law is Code (Primavera De Filippi)
Week 8
Module 4: The Rule of AI – Artificial Intelligence and Technoregulation (Part Two) 1. Technoregulation as the technological implementation of legal systems 2. Phenomenology of technoregulation: identification of protected subjects, repression of unlawful conduct, enforcement of sanctions and other forms of protection, including nudging mechanisms 3. Technoregulation “hands-on”: some experimental projects – AI4Children: identifying minors through soft biometrics – TOSware: automatic detection of abusive clauses – GigAdvisor: crowdsensing and machine learning for the protection of digital workers
Week 9
Module 5: Antigone and Artificial Intelligence – Understanding and Critically Analyzing Machine Injustice (Part One) 1. The dark sides of artificial intelligence: known discriminations and invisible implications 2. Biometric classifications and opaque forms of automated surveillance 3. Introduction to Critical Data and Algorithm Studies 4. Algorithmic injustice: definitions, emblematic cases, interpretations
Week 10
Module 5: Antigone and Artificial Intelligence – Understanding and Critically Analyzing Machine Injustice (Part Two) 1. Possibilities of resistance and algorithmic disobedience 2. Technology against technology: AI as a tool for emancipation 3. The GigAdvisor project: crowdsensing, AI and graph-based inference for critical platform evaluation 4. Open conclusions: towards algorithmic justice?
Week 11
Module 6: Regulating Artificial Intelligence – Legal Frameworks and Institutions (guest lecture) 1. The regulatory landscape of artificial intelligence: national, European and international dimensions 2. The AI Act: system classification, the notion of risk, risk categories, focus on high-risk systems 3. Ethical and legal challenges in algorithmic regulation 4. Towards a multilayered legal ecosystem for AI
Week 12
Module 7: Fair Algorithms and Online Dispute Resolution 1. Algorithmic fairness: principles, definitions, unresolved issues 2. The CREA project: computational models for predictive justice and fairness evaluation 3. Algorithms and ODR: models for online dispute resolution 4. Technologies for mediation, assisted negotiation and procedural fairness protection