2023 CBA Insolvency Law Conference: AI – What it Means for the Insolvency World

Speakers:

• Al Hounsell, Director, Strategic Innovation and Legal Design, Norton Rose Fulbright LLP (Toronto, ON)

• Maroussia Lévesque, doctoral candidate, Harvard Law School; Affiliate, Berkman Klein Center for Internet & Society; Senior Fellow, Centre for International Governance Innovation (Cambridge, MA)

• Adam Zeldin, Vice-President, Canadian Restructuring, Richter Inc. (Toronto, ON)

Moderator: Jeremy Nemers, Partner, Aird & Berlis LLP (Toronto, ON)

The panel began by explaining that artificial intelligence (“AI”) is a software service that mimics past trends and precedents to predict future outcomes, and that AI services are already widely deployed in the legal market.

The panel then explained that services such as Open AI have grown significantly in the past 12 months and will continue to improve as they obtain access to more data. It will be interesting to monitor AI’s growth in the legal world (for example CHAT GPT has improved its licensing exam results within the last twelve months going to the top of the cohort) as we continue to find possibilities to integrate these platforms in our practices.

For Lawyers (Generally):

According to the panel, if lawyers can develop a generative tree of options based on repeatable precedents, then AI can be helpful and provide initial recommendations or predicted outcomes using the preset decision tree. Specifically, lawyers can use AI to conduct the following tasks:

• AI can use structured information (preset decision tree) to make predictions regarding a likely outcome;

• Generative AI can use human entered prompts to extract information from documents (i.e.: Contracts and Proceedings) to facilitate information analysis;

• Generative AI can also be used for text generation. However, this is the riskiest of the aforementioned services, as it has been linked with the horror stories regarding the citation of non-existent cases. Notwithstanding these dangers, the panel stressed that new versions of AI are reducing these risks, which will make this function interesting to follow moving forward.

Despite these possible uses, the panel reminded attendees that AI does not think like humans. Therefore, it has technical limitations associated with its lack of emotional intelligence (i.e.: AI cannot read a room or consider emotional components of a situation). AI can’t think outside the box of past data.

Moreover, it is important to consider the implications of AI performing legal services without having met the necessary licensing requirements, particularly the ethical requirements. Lawyers must ensure that their use of AI respects the respective bar requirements regarding the illegal practice of law. The panel did specify that these considerations are currently being reviewed in the United States which should lead to clearer guidelines regarding the use of AI.

Therefore, at this stage, AI can assist lawyers but not replace lawyers.

Insolvency Perspective:

From an insolvency practitioner’s point of view, AI can provide services that are complementary to their practice, in that it can reduce the time associated with administrative tasks.

For example:

• Data Analysis: Since AI can review data much quicker than humans, and use this data to predict future outcomes, it will be possible to use AI to predict future distressed situations. This may help debtors identify and address the situation earlier than if AI was not used.

• Fraud Protection: AI can also be helpful with identifying and preventing fraud. In fact, banks have been using these technologies for several years to alert clients of potentially fraudulent transactions.

• Proofs of Claim: At this stage we can consider using AI to assist with the first stage of document review by identifying if proofs of claim forms are properly completed and contain the required supporting documents. However, this will require human oversight and review to make the final assessment of a claim.

The panel explained that professionals can think of AI similarly to a junior colleague. For example, whether a junior lawyer or AI has conducted the preliminary research or review there will always be a duty on the senior lawyer to review before anything is produced in Court.

Additionally, the panel discussed the benefits and risks of AI replacing juniors for administrative work on files. One the one hand, many associates are eager to have AI automate tasks which are more mundane in order to allow them to focus on the substantive work. On the other hand, by not having juniors conduct these administrative tasks there is a risk that they will be missing essential tools in their professional development, as it is important to understand the basics of the practices when just starting out.

According to the panel, these considerations will put into question the role of law schools, which now have to provide more practical knowledge to students so that associates can hit the ground running from day one. Moreover, students will need to learn how to interact with AI services to optimize outputs.

Future Predictions:

In the near term, the panel believes that AI is a compliment to legal work. This will create a premium for subject matter expertise and judgment, as the computer cannot make judgment calls.

Since AI can hold so much data at once, it is able to surface trends and connect the dots very quickly. It will therefore be necessary to find ways to integrate it into our practices.

For clients, AI could reduce judicial costs and improve access to justice. However, it may also lead to a two-tier system based on the affordability of lawyers. Indeed, if the value add of lawyers becomes their expertise and ability to make judgment calls, the legal system will be divided between those who can afford a lawyer and those who can only rely on AI.

Additionally, students will need to learn to work with these tools to maximize outputs. Learning how to communicate and work with AI will be a necessary skill in and of itself.

Overall, the panel predicted that the volume of legal work could increase for the following reasons:

• AI will generate a 35-45% increase in productivity in the economy in general, which will in turn generate more legal work;

• Increased access to justice means an increase in legal work. However, the courts will need to learn how to use AI to respond to the increase in demand;

• The use of AI will create substantive legal questions and a gray area in the regulation. This will generate work for lawyers who are tasked with answering these questions; and

• AI will create new alternative legal service providers which generally increase productivity, and legal work in the market.