Artificial Intelligence In English Law

1y ago
30 Views
1 Downloads
6.13 MB
49 Pages
Last View : 11d ago
Last Download : 6m ago
Upload by : Grady Mosby
Transcription

1Section titleArtificial Intelligencein English LawProfessor John Armourand Robin Dicker QCoutline an ambitiousprogramme of research intothe potential application ofAI toEnglish lawwww.southsquare.comBrexit:Deal or no Deal?The third instalmentby Mark Phillips QCon the ongoingBrexit sagaIndia’s RevisedInsolvencyFramework: Too fastfrom too slow?An update on thecurrent insolvencyregime in India, andchanges brought aboutby the Insolvency andBankruptcy Code 2016A regular review of news, cases andarticles from South Square barristers

SOUTH SQUARE DIGESTMarch 2019www.southsquare.com3In this issueIn this issue061221Artificial Intelligence in English Law:A Research AgendaBrexit: Deal or no Deal?India’s Revised Insolvency Framework:Too fast from too slow?Professor John Armour and Robin DickerQC outline an ambitious programme ofresearch into the potential applicationof AI to English lawThe third instalment by Mark PhillipsQC on the ongoing Brexit sagaCyril Shroff and Dhananjay Kumar ofCyril Amarchand Mangaldas, togetherwith South Square’s Mark Arnold QC andMatthew Abraham, provide an update onthe current insolvency regime in Indiaand the changes that have been broughtabout by the Insolvency and BankruptcyCode 2016ARTICLESBVI Strengthening28Brian Child and Matthew Freemanof Campbells, BVI, write on recentsignificant legal developmentsas the country and its businessesreturn to normal following thehurricanes and mud slides of 2017‘The set is highlyregarded internationally,with barristers regularlyappearing in courtsaround the world.’Legal Eye:Anthropology and law32Shareholder Disputes:Unfair Prejudice54David Alexander QC and AdamGoodison review the unfairprejudice legislation in the contextof shareholder disputesCompany/Insolvency set71Madeleine Jones explores what,if anything, lawyers have tolearn from anthropologistsThe South Square StoryFormer Member of ChambersSimon Mortimore QC providesus with the first instalment ofhis history of Chambers62Jeremy Goldring QC considers theCourt of Appeal’s decision in Re: OJSCInternational Bank of Azerbaijanof the year, winner 2018CHAMBERS BAR AWARDSCHAMBERS UK 44 (0)20 7696 9900 practicemanagers@southsquare.com www.southsquare.com67Gabriel Moss QC on the MergersDirective and creditor protectionSouth Square welcomes newAssociate Member ProfessorChristoph G. PaulusDead-end: permanent stay underthe Model law barred by GibbsEuroland: Creditors “Blastedwith The East Wind ”Designed and produced by Creative Interpartners, London74REGULARS3879294From the EditorNews in BriefSouth Square ChallengeDiary DatesCASE DIGESTSEditorialBanking & FinanceCivil ProcedureCommercial LitigationCompany LawCorporate InsolvencyPersonal InsolvencyProperty & TrustsSport333436384345505253South Square Digest DisclaimerThe content of the Digest is provided to you forinformation purposes only, and not for the purposeof providing legal advice. If you have a legal issue,you should consult a suitably-qualified lawyer. Thecontent of the Digest represents the views of theauthors, and may not represent the views of otherMembers of Chambers. Members of Chamberspractice as individuals and are not in partnershipwith one another.

SOUTH SQUARE DIGESTMarch 2019www.southsquare.comFrom the EditorFrom the EditorsMay now remains engaged in attemptsto re-open talks with the EU to getchanges to the “backstop”. It remainsimpossible to predict what the future ofBrexit will be.In equally unsettling news, thepost-Christmas period saw a numberof further high street insolvenciesincluding the appointment ofadministrators in relation to HMV,Oddbins and Patisserie Valerie,together with the announcement thatthe UK economy had expanded at itsslowest annual rate in six years in 2018.William Willson and Marcus HaywoodWelcome to first edition of theSouth Square Digest for 2019The four months since the last editionof the Digest was published in October2018 have seen the handing down ofjudgments in a number of importantcases in which members of chambershave been involved, including Re OJSCInternational Bank of Azerbaijan (wherethe long standing rule in Gibbs and itsinterrelationship with the Cross BorderInsolvency Regulations 2006 wasconsidered by the Court of Appeal), ReNoble Group Limited (concerning a highlycomplicated restructuring, involvingan English scheme of arrangement ofone of the world’s biggest commoditytraders), Lehman Brothers Australia vLomas (where Hildyard J considered thescope of the rule in ex parte James) andBresco Electrical Services Ltd v MichaelJ Lonsdale (Electrical) Ltd (where theCourt of Appeal gave guidance onthe interplay between constructionadjudication and insolvency regimes).In wider news, Brexit continues todominate the headlines. As at thetime of writing, “exit day” remains 29March 2019. However, there remainsno consensus. On 15 January 2019, anoverwhelming and decisive majorityof MPs rejected the WithdrawalAgreement negotiated between the EUand the UK government. Then, on 29January 2019, with just 59 days to gountil “exit day”, MPs narrowly passeda government-backed amendment,tabled by Graham Brady MP, proposingthe replacement of the Irishborder “backstop” with unspecified“alternative arrangements”. TheresaAgainst this background of uncertainty,this edition of the Digest contains anumber of topical articles. ProfessorJohn Armour of the University ofOxford and Robin Dicker QC considerthe topic of artificial intelligence(AI) in English law. AI, once a notionconfined to science fiction novels,movies and research papers, is nowmaking a tremendous impact onsociety. Whether we are aware of itor not, AI already pervades much ofour world, from its use in bankingand finance to electronic disclosure inlarge scale litigation. The applicationof AI to English law raises manyinteresting questions, a number ofwhich will be explored by a programmeof research being undertaken by aninterdisciplinary team of academics atOxford, as John and Robin explain.Ever topical, Mark Phillips QCcontinues his Brexit series withan article which discusses thepossible frameworks for crossborder insolvencies and schemes ofarrangement following the UK’s exitfrom the EU.Following its recovery from HurricaneIrma, Brian Child and MatthewFreeman of Campbells, review somerecent developments in the BritishVirgin Islands’ legal market.Closer to home, in the first of a seriesof articles by Simon Mortimore QCtracing the history of South Squarefrom its origins to the present day,Simon provides a lively and fascinatingaccount of the early career of CyrilSalmon KC and the beginnings ofchambers.In his regular “Euroland” piece, GabrielMoss QC reflects on the judgmentof Snowden J in Re M2 PropertyInvest Limited which considered theinterrelationship between creditorprotection and the EU Directive oncross-border mergers. David AlexanderQC and Adam Goodison consider theunfair prejudice legislation and recentcase law on shareholder disputes. AndJeremy Goldring QC considers theCourt of Appeal’s judgment in Re OJSCInternational Bank of Azerbaijan.Many thanks to all for theircontributions. As always, viewsexpressed by individual authors andcontributors are theirs alone.We hope you enjoy this edition ofthe Digest. And if you find yourselfreading someone else’s copy and wishto be added to the circulation list,please send an email to kirstendent@southsquare.com and we will do ourbest to make sure that you will get thenext edition and all future editions.Marcus Haywood and William WillsonFor an alternative window onto thelegal world, Madeline Jones’ “Legal Eye”turns to the topic of anthropology andthe law and asks “Are you a rainmaker?”Finally, we have the ever-popular SouthSquare Challenge, which for this editionchallenges you to match the judge withthe correct hobby It goes without saying that if you haveany feedback to give us in relation tothe Digest – positive or negative – wewould be delighted to hear from you.Meanwhile, Cyril Shroff and DhananjayKumar of Cyril Amarchand Mangaldastogether with Mark Arnold QCand Matthew Abraham considerIndia’s recently revised insolvencyframework and the changes thathave been brought about in India bythe Insolvency and Bankruptcy Code2016, a landmark event for the Indianinsolvency regime.CherryX per Wikimedia Commons International Bank Azerbaijan5

SOUTH SQUARE DIGESTMarch 2019www.southsquare.comArtificialIntelligencein EnglishLaw:A ResearchAgendaArtificial intelligence (AI)is attracting an enormousamount of attention in themedia and public discourse.Well-publicised recent successesfor AI have included self-drivingcars and self-teaching boardgame champions.Economists see AI as a nascent general purposetechnology, capable of transforming workingpatterns in professional sectors, including law, in away that some liken to the impact of the industrialrevolution on manual labour.1 This disruption canbring great efficiencies, but also displace manyhuman employees.2Legal services are a major contributor to the UKeconomy, accounting in 2016 for 1.5% of domesticGVA and generating a trade surplus of 4bn.3 Ifimplemented effectively, AI offers opportunitiesto improve legal services both for commercialparties and individuals. We are collaborating,along with an interdisciplinary team of academicsat Oxford and a range of other private sectorpartners, on an ambitious programme of researchinto the potential application and limitations ofAI to English law. The project, entitled Unlockingthe Potential of AI for English Law,4 is funded byan award from UK Research and Innovation aspart of its Next Generation Services investmentprogramme.5 This in turn is one of a number ofIndustrial Strategy Challenge Funds, establishedto stimulate research partnerships betweenacademia and the private sector in areas ofimportance to the UK economy.The research project will investigate severalof the (many) important questions raised forlaw and lawyers by the advent of AI. How isAI being used in legal services, and how doesorganisational structure and governance affectits implementation? What are the possibilitiesfor the adoption of AI in dispute resolution?Will lower costs facilitate access to justice? Willthere be an impact on quality of provision thatmeans the “justice” thereby provided is lackingin one or more important respects? What tradeoff, if any, should be adopted between cost andquality? What constitutional and other constraintsare there on the use of AI in legal proceedings?Are there technological advances in the pipelinethat may further push back the boundarybetween humans and machines in the future?How is AI adoption affecting job descriptions?What implications are there for the boundariesof professional knowledge, business models inlegal services, and the education and trainingof lawyers and relevant technical specialists?In this article, we provide some background tothese research questions along with some verypreliminary insights from our work.What is AI?PROFESSORJOHN ARMOUR,OXFORD UNIVERSITYROBIN DICKER QC7AI in English LawAI is not a new concept, the first usage generallybeing attributed to computer scientists JohnMcCarthy and Myron Minsky in the mid-50s.6 Theconvention is to use the term in a functional sense,meaning that an artificial system functions as wellas, or better, than a human. Clearly, machines canperform many tasks better than humans that donot involve intelligence, as opposed to strength orendurance. The “intelligence” qualifier thereforecan usefully be understood as restricting thecomparison to activities for which a human woulduse their brain – most obviously, processingand analysing information.The classic assessment of whether a systemfunctions as well as a human is the so-called“Turing test”, in which a human is asked toengage in conversation with messages sentthrough a mechanism that does not revealwhether the party on the other side is humanor not.7 If a human participant cannot distinguishthe communications of an artificial system froma human, then the test is passed by that system.To pass a Turing test without any constraintsaround the type of conversation that could behad, the machine would need to exhibit artificialgeneral intelligence (AGI); that is, as good ashuman in every dimension of intelligence.Modern AI systems do not come anywhere nearAGI. This is—according to experts—anywherebetween a decade and two centuries away.8Rather, the AI deployed today only has (super)human-level capability in respect of narrowlydefined functions, such as image recognition,driving vehicles in straightforward surroundings,or the classification of documents.We are collaborating, alongwith an interdisciplinary teamof academics at Oxford and arange of other private sectorpartners, on an ambitiousprogramme of research intothe potential applicationand limitations of AIto English law1. See e.g. EBrynjolfsson, D Rock,and C Syverson,‘Artificial Intelligenceand the ModernProductivity Paradox:A Clash of Expectationsand Statistics’ and MTrajtenberg, ‘AI as theNext GPT: A PoliticalEconomy Perspective’,in AK Agrawal, J Gansand A Goldfarb (eds.),The Economics of ArtificialIntelligence: An Agenda(Chicago: Universityof Chicago Press,forthcoming 2019).2. See e.g. J Furmanand R Seamans, ‘AI andthe Economy’ in J Lernerand S Stern, InnovationPolicy and the Economy2018, Vol 19 (Chicago:University of ChicagoPress, forthcoming2019).3. TheCityUK,UK Legal Services2017: Legal Excellence,InternationallyRenowned (London:TheCityUK, 2017).4. See h-law/workpackages.5. ychallenge-fund/nextgeneration-services/.6. J McCarthy, MLMinsky, N Rochesterand CE Shannon,A Proposal for theDartmouth SummerResearch Project onArtificial Intelligence(1955) (proposing a“2 month, 10 manstudy of artificialintelligence”,organised around“the conjecture thatevery aspect oflearning or any otherfeature of intelligencecan in principle beso precisely describedthat a machine can bemade to simulate it.”).7. AM Turing,‘Computing Machineryand Intelligence’ (1950)49 Mind 433, 434. A testso formulated “has theadvantage of drawinga fairly sharp linebetween the physicaland the intellectualcapacities of a man.”(ibid).8. Martin Ford,Architects ofIntelligence(Birmingham:Pakt Publishing,2018), 528-9.

SOUTH SQUARE DIGESTMarch 2019www.southsquare.com9AI in English LawFEATURE ARTICLE: ARTIFICIAL INTELLIGENCE IN ENGLISH LAWThere havebeen at leastthree distincttechnicalapproachesto AI sincethe birth ofthe field99. See S Russell andP Norvig, ArtificialIntelligence: A ModernApproach, 3rd ed. (UpperSaddle River, NJ: Pearson,2010), 16-28.10 See generally, JMinker, ‘Introduction toLogic-Based ArtificialIntelligence’, in J Minker(ed.), Logic-Based ArtificialIntelligence (Dordrecht:Kluwer, 2000), 3.11 See e.g., P Harmonand D King, ExpertSystems: ArtificialIntelligence in Business(New York: Wiley, 1985).For a classic accountof their application tolaw, see R Susskind,Expert Systems in Law: AJurisprudential Enquiry(Oxford: ClarendonPress, 1987).12 See A Halevy, PNorvig and F Pereira,‘The UnreasonableEffectiveness of Data’(2009) IEEE IntelligentSystems 8.13 See e.g., F Chollet,Deep Learning with Python(Shelter Island, NY:Manning, 2018), 8-11.14 See Ford,supra n 8, 186.Defining AI in this functional way means thatno particular restrictions are put on the natureof the computing system used. Indeed, there havebeen at least three distinct technical approachesto AI since the birth of the field.9 The first, popularin the 1960s, involved logical rules: the idea wasto develop a general-purpose system capable ofderiving answers to problems through formallogical reasoning.10 This approach fell afoul ofthe problem that deterministic calculation ofoutcomes becomes exceedingly complex for evenmoderately challenging real-world problems.A different tack was taken in the 1980s and 90s,with the advent of so-called “expert systems”.11These were designed to give human users thebenefit of expert answers to problems in an areaof practice. The problems and answers – whatcomputer scientists call “domain knowledge” –were characterised with the help of relevanthuman experts. These were then coded into asystem designed to answer questions relatedto that particular body of knowledge. Expertsystems in turn proved quite brittle, however.If the question asked by a user fell outside thesystem’s expertise, it could not give an answer.And moreover, the framing of questions hadto be done in terms of the specific syntax ofthe system. If the user was unable to expressthemselves in terms the system could understand,then again it would fail. These created roadblocksto the roll-out of such systems.Recent advances in AI rely primarily on machinelearning (ML). This is an approach to computingin which the solution to an optimisation problemis not coded in advance, but is derived inductivelyby reference to data. The technique relies onapplying computing power to very large amountsof data, the availability of which has blossomedin recent years.12 Progress since 2012 has largelybeen in a particular type of ML known as deeplearning, which involves running multiple layersof representation of the data in series.13The greatest practical successes with ML to datehave been in the use of supervised learningtechniques.14 This refers to a process that beginswith a dataset that is classified or labelled byhumans according to the dimension of interest,known as the training data. The system analysesthis dataset and determines the best way topredict the relevant outcome variable (classifiedby the experts) by reference to the other availablefeatures of the data. The nature of the features,and the relationships between them, relevantfor predicting the outcomes can be exceedinglycomplex: the power of ML lies in identifyingthe optimal mix of input variables. The trainedmodel—that is, the algorithm with the set ofparameters that optimised performance on thetraining dataset—is then put to work on a newtest dataset, to see how effective it is at predictingoutside the original training sample. These resultsmust now be checked by human experts.3. Applying AI to Law: EstablishedApplicationsOne of our research questions is to understandthe way in which AI is currently being applied inlegal services. The foregoing account reveals twokey constraints on the application of ML-basedAI to legal contexts. First, the need for a large andrelevantly labelled dataset for training the model.And second, the need for consistency between thetraining dataset and the data on which the trainedmodel is to be used for predictive purposes.Supervised learning techniques have beenenormously effective in image recognitionand language translation contexts, where vastquantities of pre-labelled data are available onthe internet, and there is high consistency in theformat of data. In many legal contexts, however,these conditions may prove more restrictive.Labelling a sufficiently large dataset of legaldocuments is costly, and the more varied thedocument types in question, the more difficult itmay be to get good results. This means supervisedlearning techniques have fixed costs to implement,and their generalisability is constrained. They areconsequently most useful in contexts where thereis a very high volume of very similar material.3.1 Technology-assisted reviewThe contexts in which ML-based techniques arenow being actively applied in legal services areto identify relevant documents from amongstvery large bodies of materials. In contentiousmatters, this is known as “technology-assistedreview” (TAR). The growth of electronically storedinformation (ESI) means that there are enormousvolumes of potentially relevant information fordiscovery/disclosure in a typical contentiousmatter. A large contentious matter can easilyrequire review of hundreds of gigabytes of ESI perparty.15 This has triggered a rapid rise in the costsof pretrial discovery relative to overall liti

the topic of artificial intelligence (AI) in English law. AI, once a notion confined to science fiction novels, movies and research papers, is now making a tremendous impact on society. Whether we are aware of it or not, AI already pervades much of our world, from its use in banking and finance to electronic disclosure in large scale litigation. The application of AI to English law raises many .

Related Documents:

and artificial intelligence expert, joined Ernst & Young as the person in charge of its global innovative artificial intelligence team. In recent years, many countries have been competing to carry out research and application of artificial intelli-gence, and the call for he use of artificial

BCS Foundation Certificate in Artificial Intelligence V1.1 Oct 2020 Syllabus Learning Objectives 1. Ethical and Sustainable Human and Artificial Intelligence (20%) Candidates will be able to: 1.1. Recall the general definition of Human and Artificial Intelligence (AI). 1.1.1. Describe the concept of intelligent agents. 1.1.2. Describe a modern .

IN ARTIFICIAL INTELLIGENCE Stuart Russell and Peter Norvig, Editors FORSYTH & PONCE Computer Vision: A Modern Approach GRAHAM ANSI Common Lisp JURAFSKY & MARTIN Speech and Language Processing, 2nd ed. NEAPOLITAN Learning Bayesian Networks RUSSELL & NORVIG Artificial Intelligence: A Modern Approach, 3rd ed. Artificial Intelligence A Modern Approach Third Edition Stuart J. Russell and Peter .

Peter Norvig Prentice Hall, 2003 This is the book that ties in most closely with the module Artificial Intelligence (2nd ed.) Elaine Rich & Kevin Knight McGraw Hill, 1991 Quite old now, but still a good second book Artificial Intelligence: A New Synthesis Nils Nilsson Morgan Kaufmann, 1998 A good modern book Artificial Intelligence (3rd ed.) Patrick Winston Addison Wesley, 1992 A classic, but .

BCS Essentials Certificate in Artificial Intelligence Syllabus V1.0 BCS 2018 Page 10 of 16 Recommended Reading List Artificial Intelligence and Consciousness Title Artificial Intelligence, A Modern Approach, 3rd Edition Author Stuart Russell and Peter Norvig, Publication Date 2016, ISBN 10 1292153962

PA R T 1 Introduction to Artificial Intelligence 1 Chapter 1 A Brief History of Artificial Intelligence 3 1.1 Introduction 3 1.2 What Is Artificial Intelligence? 4 1.3 Strong Methods and Weak Methods 5 1.4 From Aristotle to Babbage 6 1.5 Alan Turing and the 1950s 7 1.6 The 1960s to the 1990s 9 1.7 Philosophy 10 1.8 Linguistics 11

Artificial Intelligence, Machine Learning, and Deep Learning (AI/ML/DL) F(x) Deep Learning Artificial Intelligence Machine Learning Artificial Intelligence Technique where computer can mimic human behavior Machine Learning Subset of AI techniques which use algorithms to enable machines to learn from data Deep Learning

Artificial Intelligence and Its Military Implications China Arms Control and Disarmament Association July 2019 What Is Artificial Intelligence? Artificial intelligence (AI) refers to the research and development of the theories, methods, technologies, and application systems for