Graduate Diploma in Legal Professional Practice (GDLPP)

PLTX2040 Practicum (Technology and Innovation)

Readings on AI and on ‘AI & Law’

Graham Greenleaf & Philip Chung
30 June 2020

This reading guide is to assist students who wish to obtain a deeper self-directed understanding of both the general field(s) of AI, and specific applications of AI to law. Readings are primarily online, but a number of books are included.

In the week before the first class, students should at least explore readings on AI generally, and if time permits, start reading on ‘AI and law’. Use of the readings should then continue throughout the course, in order to obtain competence in understanding developments in AI and law, and in the general context of AI development in which it occurs.

Artificial intelligence (AI) generally

Survey articles and books

• Melanie Mitchell Artificial Intelligence – A Guide for Thinking Humans (Pelican2019) –This introduction is relatively non-technical, and assesses what evidence there is (or is not) for the success/effectiveness of particular branches of AI.
• Margaret Boden Artificial Intelligence – A Very Short Introduction (OUP 2016) This introduction is also very readable, but is perhaps less sceptical.
• Wikipedia: Artificial intelligence – A rich introduction to the history and varieties of AI, with links to many other Wikipedia entries, and references. As good a starting point as can be found in most books.
• M. Legg and F. Bell Artificial Intelligence and the Legal Profession - A Primer (flipstream, UNSW Faculty of Law, 2018) – see ‘What is artificial intelligence’ (pgs. 2-12). This is valuable because it is written with a legal audience in mind.
• S Russell and P Norvig Artificial Intelligence: A Modern Approach (3^rd Ed) (Pearson, 2016) This is the standard student textbook on AI; highly technical and comprehensive; Chapter 1 is a more readable introduction to the field, and the concluding chapter is also non-technical.

Specific AI technologies and branches

Each of the general introductions above has sections on each of the specific technologies below.

Symbolic AI

• Wikipedia: Symbolic artificial intelligence – Too brief to be much use.
• Ben Dickson ‘What is symbolic artificial intelligence?’ (TechTalks, 2019) - Journalistic
• Richa Batia ‘Understanding the difference between Symbolic AI & Non Symbolic AI’ (2017) – Another journalistic approach.

Neural networks

• Pathmind: A Beginner's Guide to Neural Networks and Deep Learning – Thestarts as a step-by-step introduction to neural networks, with examples, but by the end is quite complex.
• Wikipedia: Artificial neural networks – [Difficult, somewhat confusing introduction]

‘AI & Law’

Surveys and introductions

• Kevin Ashley Artificial Intelligence and Legal Analytics (Cambridge, 2017) – This is the only textbook survey of the field of AI and law from an academic perspective; often technical, and with an emphasis on case-based reasoning (Ashley’s speciality), but with good coverage of legislation-based systems as well.
• M. Legg and F. Bell Artificial Intelligence and the Legal Profession - A Primer (flipstream, UNSW Faculty of Law, 2018) – see ‘How is AI used in legal services’ (pgs. 13-20).
• Wikipedia: Legal informatics: Artificial intelligence – Introduction to the history of AI & law, including conferences, journals, and main research topics. Use this to explore other Wikipedia pages on AI & law, such as:
  • Wikipedia: Lawbot – ‘Lawbots are a broad class of customer-facing legal AI applications that are used to automate specific legal tasks’. Has some examples of systems in practice.
  • Wikipedia: Legal expert system – [Grain of salt needed] Shows very limited knowledge of history of the field.

‘Law as Code’ groups

• aka ‘Rules as Code’, ‘Legislation as Code, Digital Legislation, RegTech
• OECD Observatory of Public Sector Innovation (OPSI)
  • Draft ‘Rules as Code’ primer Cracking the Code: Rulemaking for humans and machines (OECD, May 2020) – This very lengthy (129 pgs) primer on ‘Rules as Code’ (RaC) is a good introduction to the field, and has a lengthy bibliography.
  • James Mohun and Alex Roberts Seeking your feedback on draft Rules as Code primer, 27 May 2020
  • AustLII(Greenleaf, Mowbray, Chung) Strengthening development of Rules as Code: Submission to the OECD’s OPSI on Cracking the Code, 23 June 2020
• Tim de Sousa (and others) Rules as Code Handbook (wiki)
  • 7.1 Rules as Code projects around the world
  • Rules as Code (NSW Govt) – ‘Policy Lab’ is part of NSW government
    • Announcement of NSW ‘Rules as Code’ initiative (Jan 2019, Tim de Sousa, Director, Policy Lab); Progress report, May 2019
    • ‘Emerging Technology Guide: Rules as Code’ - many valuable links from here
    • Using the (French) ‘rules platform’ OpenFisca.
    • ‘Working with the Better Regulation Division to code parts of the new Community Gaming Regulation’
    • See PPTs and video from ‘Rules as Code Showcase’ (2019)
  • New Zealand developments
    • ‘SmartStart’ NZ government Open Fisca app
    • Digital.Govt.NZ NZ LabPlus: Better Rules for Government Discovery Report Summary and full report (April 2018)
    • ‘Law as Code in NZ’ – Law Foundation investigation, 2020
  • Data61’s ‘Regulation-as-a-Platform’
    • Data61 is the data science part of the CSIRO; it is involved in both government and business applications.
    • ‘Business processes and legal informatics’ blurb. All links now go to this.
    • “Case study on CSIRO’s Data61, Australia: Contribution to the OECD TIP Digital and Open Innovation project” (2019) – Background on Data61; see pp. 30-31 for ‘Regulation as a Platform’
    • Regorous (2015) ‘The underlying technology is Defeasible Deontic Logic which maps rules and regulations, as well as a company’s business processes, into equations.’
  • Digital Transformation Agency (DTA - Cth) Digital Legislation Working Group
    • Multi-jurisdictional; No meetings since early 2019?
  • Statutory draftsperson’s perspective
    • Matthew Waddington, (2019), ‘Machine-consumable legislation: A legislative drafter’s perspective – human v artificial intelligence’, The Loophole, no. 2, pgs. 21-53.

‘Business process modelling’

• To some extent, this is ‘law as code’ ideas applied to the private sector
• Wikipedia: Business Process Model and Notation (BPMN) – formal notation
• Business Rules Community (includes Business Rules Journal)

Academic research on AI & Law

• Artificial Intelligence and Law (Journal) – plus IAAIL and ICAIL
  • The leading academic journal since 1991, edited by members of the International Association of AI & Law (IAAIL), which organises the biennial International Conference on AI and Law (ICAIL)
  • The Journal has 37 open access articles.
• JURIX (The Foundation for Legal Knowledge-based Systems)
  • Netherlands-based foundation; JURIX annual conference (in 32^nd year 2019) was originally based in the Netherlands, but is now held in many European locations.
  • Papers from JURIX conferences are often published by IOS Press for open access – eg Araszkiewicz, & Rodríguez-Doncel (Eds) Legal Knowledge and Information Systems (2019)
• JURISIN (Juris-Informatics)
  • This Japanese conference with an international orientation held its 13^th Conference in 2019.
  • Selected papers are published as a post-proceedings via Springer Verlag LNAI (Lecture Notes in Artificial Intelligence). Some can be found for free access.
• Reviews of the field
  • T Bench-Capon et al ‘A history of AI and Law in 50 papers: 25 years of the international conference on AI and Law’(2012) Artificial Intelligence and Law volume 20, pages215–319. (Not free access – but bibliography is).
  • P. Leith ‘The rise and fall of the legal expert system’ European Journal of Law and Technology (EJLT) Vol. 1 No. 1 (2010): Special Issue on Legal Informatics
• DataLex Project papers
  • Mowbray, Chung, Greenleaf, ‘Utilising AI in the Legal Assistance Sector – Testing a Role for Legal Information Institutes’ LegalAIIA ’19, June 17, 2019, Montréal; now published in (2020) Computer Law & Security Review
  • Greenleaf, Mowbray, Chung ‘Building sustainable free legal advisory systems: Experiences from the history of AI & law' (2018) 34(1) Computer Law & Security Review