Riding the GenAI wave

Artificial intelligence seems to surround us like an ocean. It anticipates our words in email, finds the fastest routes in online maps, presents us with targeted content in our social media feeds, and is implicated in recent dramatic failures at the courts. It is also quickly disrupting the way lawyers practice and assess legal risk.

Generative AI is an emerging technology. Typically, emerging technologies move towards wider adoption following a pattern called a hype cycle. It begins with excitement and high expectations for the technology. Disillusionment follows, interest wanes and some iterations of the technology may fail. Improvements are then made, and mainstream adoption may begin in earnest.

GenAI and its application to legal technology is likely still in the early phases of the hype cycle, with varying levels of comfort and risk tolerance among lawyers. Some are already on the wave's crest, testing the technology's limits. Others are in the water, paddling carefully. And some are watching from the beach with their shoes firmly still on.

Regardless, when it comes to GenAI, all lawyers should stay informed and understand how the technology may influence or alter their legal practice so that they can avoid hyper inflated expectations and provide the best possible advocacy before the courts and advice to their clients.

It is important to understand the difference between AI, machine learning and GenAI. They are not identical. Some legal tools integrate GenAI, while others simply employ AI and machine learning programs. A product that employs machine learning is not necessarily using GenAI.

Broadly, AI is the study and development of computer systems that can perform tasks that normally require some level of human intelligence. Some of these computer systems use machine learning – statistical algorithms that generalize and infer patterns from training data.

Machine learning has been used in legal technology for some time, for example in programs that check grammar, whittle down discovery production to relevant and material records, or locate and extract a particular clause in an agreement.

GenAI chatbots, such as ChatGPT, are built on large language models that have been trained on enormous amounts of data – the equivalent of tens or hundreds of millions of books – and use computational resources that cost upwards of tens of millions of US dollars. They can generate text in response to prompts using algorithms that predict the next most likely words in a sequence, so it appears that they are producing complete and plausible sentences one word at a time.

Ethically, lawyers must understand and use technology in their practice. Doing so satisfies the obligations articulated in the Federation of Law Societies of Canada's Model Code of Professional Conduct to perform legal services undertaken on a client's behalf to the standard of a competent lawyer, and to know about and use technology to maintain the required level of competence (Rule 3.1-2).

Being technologically competent does not require lawyers to adopt technology for the sake of doing so. Instead, where appropriate, a lawyer should consider what technology is relevant to their practice and use it with a solid understanding of its benefits and risks, particularly with respect to confidentiality. For example, using GenAI to research the law but not reading the cases that the program generates to confirm that they are on point and not hallucinations, would fail to meet the standard. Indeed, at present, using GenAI to draft written submissions fails to meet professional obligations and may breach court-issued directives.

While ethical rules, including the Model Code, are clear that not every tool is necessary for every lawyer’s practice, even the most technophobic practitioner should understand what AI and GenAI are and how they might benefit (or be a risk to) their clients as the technology gains momentum in the legal world.

Lawyers must be aware of the strict parameters placed around the use of GenAI by certain law societies and courts. Litigators must also consider the significant practical risks in failing to understand how technology and GenAI works as it is now integral to providing quality legal advice.

A litigator's reputation may be affected by failing to properly use technology. Recall the unfortunate lawyer who struggled to remove a filter from his face during a virtual court appearance, spawning the now infamous "lawyer cat" meme. In more serious misadventures, New York lawyers in two separate matters cited cases that ChatGPT had hallucinated. In one of those matters, the lawyers were sanctioned by the Court. Recently, similar issues have occurred in British Columbia.

It is crucial that practitioners understand the benefits and pitfalls of GenAI as it evolves because clients, opposing counsel, students, and junior associates may already be using the technology in their research and legal submissions. To identify when GenAI has been employed and to fix any issues that may arise, the skilled practitioner must have a basic understanding of how GenAI works and what it can, cannot, and should not do.

To reach the wave from the beach, reticent practitioners should not be afraid to take off their shoes and step into the water. While there is no need to dive in headfirst, you should know what GenAI is and what it does. Read the available guidance from the courts and law societies and monitor the available news about GenAI. Try out a chatbot like ChatGPT to test its capabilities and note its failings.

As for the practitioners who are already in the water and paddling, look where you want to go. Talk to law librarians, technologists, or other knowledgeable practitioners who are willing to point you in the right direction. Find out how GenAI is being incorporated into existing legal technology. Join a pilot program to try out new technology or attend a conference to learn more.

Practitioners who are near the crest of the wave should consider using their perspective to guide others and avoid hazards. Share your knowledge and experience with your colleagues at your organization. Volunteer to lead a pilot to deepen your expertise, join a steering committee that is deciding GenAI best practices, or teach a CLE session.