Meanwhile, Canada Also Launches Major Legal AI + Justice Research Project

In one of those unexpected moments of synchronicity, Canada has also just launched a major research project into AI and justice that, like the UK’s commission on algorithms and justice that was launched yesterday by the Law Society, may also produce recommendations on ethics and regulation of the technology, with a central aim to increase access to justice through the use of AI.

Canada’s Cyberjustice Laboratory, based in Montreal, announced last night the launch of the ACT (Autonomy through Cyberjustice Technologies) research partnership, led by Professor Karim Benyekhlef, Director of the Laboratory, and funded by the Social Sciences and Humanities Research Council of Canada’s (SSHRC) Partnership Grants programme.

Created in 2010 by professors Benyekhlef (University of Montreal) and Fabien Gélinas(McGill University), the Cyberjustice Laboratory aims to reimagine the delivery of justice. Over the next six years, ACT will work with an international team of 45 researchers, as well as 42 partners including research centres, public institutions, legal professionals, representatives from civil society and private sector actors.

ACT will also get a CAN$2.5 million grant from the SSHRC, as well as CAN$4.3 million in contributions from its various partners (or CAN$6.8m, which is around US$5.17m).

Work will focus on the design and simulation of AI tools for conflict prevention and resolution. The team will also seek to propose best practice guidelines and a governance framework that should allow decision makers to develop informed public policies based on high quality scientific work…..which in fact seems to be very similar to what the UK has just started…..which makes Artificial Lawyer wonder if the two groups should perhaps work together on this….?

This major project will also foster the next generation of multidisciplinary researchers in law and computer science by contributing to the specialised training of approximately 25 students annually. In terms of secondary benefits, the research sub-projects will experiment with and implement innovative technology for justice, and will also provide new opportunities for the private sector, said the organisation.

As there suddenly seems to be so much interest now in this area on both sides of the Atlantic, below are listed the main areas of research that ACT will conduct as it may provide inspiration to others around the world. The 16 areas are:

Working Group 1 – Conflict prevention

Supbroject 1

Pre-conflict decision tools
The aim is to develop tools to assist decision-making prior to conflict. This includes any settlement of the dispute prior to its being brought to court. This field includes instruments freely accessible to all citizens to enable them to determine the extent of their rights.

Subproject 2

Pre-conflict decision tools for legal practitioners
The objective of this project is to analyze the contribution of algorithmic tools capable of predicting the probable outcome of a trial. It is important to understand how these models work in order to justify their reliability and to be able to make them a useful working tool for practitioners.

Subproject 3

Decision support tools for authorities and the police
The specificity of prosecuting authorities is linked to the public interest at stake in their decision to take legal action. It is important for them to analyse in a more global way all the data relating to their scope of intervention in order to target the most effective and dissuasive behaviours.

Subproject 4

Smart contracts and regulatory technologies. The development of technological contracts involving the substitution of the third party by computer technology is making significant progress, requiring changes in state legislation. The future of contractual obligations involves an analysis by the legal community to facilitate their proper implementation.

Working Group 2 – Conflict resolution

Subproject 5

Online dispute resolution: pilot project with the Office de la protection du consommateur du Québec (OPC)
Since the fall of 2016, the Online Dispute Resolution Support Platform developed by the Cyberjustice Laboratory (PARLe), has been providing consumers and merchants in Quebec with a fast and free service to resolve their disputes. Important lessons need to be learned from this experience in order to understand the factors determining the rate of out-of-court settlement of disputes and to analyse the data from the perspective of the consumer, the merchant, the mediator, the Laboratory registry and the Consumer Protection Office (OPC).

Subproject 6

Online dispute resolution: pilot project with the Condominium Authority Tribunal (CAT)
The Condominium Authority Tribunal launched an online dispute resolution platform for condominium disputes in November 2017. A Canadian first, this 100% virtual forum provides negotiation, mediation and adjudication services without any physical courtroom. It will be necessary to take stock of this experience, the aim of which is to reduce the costs of justice and increase its accessibility to as many people as possible.

Subproject 7

Distributive and immersive justice. The implementation of technological tools in the context of the conduct of hearings (e. g. video appearance) modifies the modes of interaction between judicial actors. It is therefore necessary to examine the progress being made in this area and to analyse the social, cultural and psychological effects of these transformations in order to determine whether they improve or worsen the sense of procedural fairness.

Subproject 8

Tools for self-represented litigants. The phenomenon of self-representation creates a risk of congestion in the courts for cases that are inadmissible, have low stakes, or are unlikely to succeed. In order to overcome these difficulties, many tools are being developed in the field of citizens’ self-representation and must therefore be evaluated in terms of their effectiveness, relevance and satisfaction rates of the target audience.

Subproject 9

Decision tools for arbitrators, judges and adjudicators. The development of decision-making tools is the subject of several abstract models. Both their experimental implementation and their practical evaluation run up against the reluctance of those who are competent to make legal decisions. Transparency and pedagogical efforts are needed to ensure the interest and reliability of these tools, which involves examining already existing decision-support techniques.

Working Group 3 – Governance and Policy

Subproject 10

Inventory of policies and best practices. Advances in cyberjustice depend on local initiatives that are proven to be successful where they are initiated. It is important to participate in this effort to disseminate good practices by identifying projects and perspectives that are taking cyberjustice one step further in Canada, the United States and Australia.

Subproject 11

Protecting vulnerable groups. The development of judicial technologies creates the risk of neglecting the vulnerabilities of certain individuals, a notion closely linked to that human empathy. It is therefore necessary to identify the tensions generated by these tools among vulnerable populations. Initiatives aimed at integrating ethics with artificial intelligence could make it possible to better develop these tools in their cultural dimension.

Subproject 12

Evaluation, key performance and quality indicators, and best practices. Measuring the progress of cyberjustice requires reliable quality and performance indicators that are supported by the scientific community. Given the importance of convincing judicial actors of the benefits of cyberjustice, methods of evaluating tools are important, including from the perspective of international implementation in different legal systems, such as in continental Europe and common law countries.

Subproject 13

Ethical issues with the use of algorithms for justice. The use of algorithms for justice implies the collection of data that enriches the potential of these tools with no predictable certainty as to the type of advancement available tomorrow. Ethics is fundamental to this mechanism, since advances in these technologies must respect fundamental human rights, which requires the incorporation of benchmarks to ensure that these tools are transparent and comply with the criteria of a democratic society.

Subproject 14

Issues associated with the valorisation of judicial data: intellectual property, privacy and open data. Data collected in the judicial field are sensitive and often need to be anonymised. Access to these data must be subject to regulatory mechanisms that clarify what is to be deleted, what can be used and exchanged, under what conditions and for what purposes.

Subproject 15

Security issues. Hacking tends to become an increasingly worrying threat as computer technology develops. The protection of the digital environment is therefore essential for its development, which requires the creation of cybersecurity conditions for judicial actors.

Subproject 16

Roadmap for digital transformation. There are problems common to all the difficulties associated with cyberjustice. Feedback from each stakeholder should be construed as useful information for all. A general and methodical approach must therefore be conceptualized in order to measure the level of transformation already achieved and to be achieved for judicial actors.

So, there you go. Looks like there are a lot of good ideas here that we can all consider in relation to the development of regulation of legal AI systems. The only question Artificial Lawyer has now is: which other country is about to launch a massive research project into law and AI? If you are about to launch one, or are thinking of doing so, please drop me a line. And to find out more on what’s happening in Canada, check out the main ACT page here. To find out more about the UK’s own project, check out the story here.

1 Trackback / Pingback

  1. Meanwhile, Canada Also Launches Major Legal AI + Justice Research Project | Artificial Lawyer – Collective Intelligence

Comments are closed.