news + views + events
Balancing Innovation + Integrity: Navigating AI's Impact on the Legal Profession

A Cautionary Tale

In a recent and notable incident in British Columbia, the legal profession was given a stark reminder of the pitfalls associated with the use of Artificial Intelligence (AI) in legal practice. A Vancouver lawyer faced repercussions for citing two nonexistent cases generated by ChatGPT in court documents. This misuse of AI technology, albeit without intent to deceive, highlighted the critical importance of diligence and the potential dangers of over-reliance on AI for legal arguments. The lawyer was ordered to compensate for the opposing counsel's efforts to verify the non-existent caselaw, underscoring the judiciary's concerns over AI-generated content's integrity in legal documentation.

National Response: Disclosure + Verification Requirements

As the legal profession grapples with the integration of AI, the Federal Court of Canada, along with jurisdictions such as Manitoba, Nova Scotia, and the Yukon, is setting a precedent in transparency and ethical use of AI in legal proceedings. These bodies issued notices that mandate that all litigation-related documents containing AI-generated content must explicitly declare this use in their opening paragraphs. Alberta Courts issued a similar notice voicing concerns over AI, but have not yet said that AI’s use must be disclosed in litigation-related documents.

These measures reflect a broader understanding of the challenges and opportunities presented by emerging technologies. The Federal Court, in particular, has voiced concerns about the potential for "deepfakes", AI-generated fabrications of legal authorities, and the use of generative decision-making tools by officials. To address these issues, it emphasizes the necessity for lawyers to offer traditional, human-centric services, especially when representing clients who may not be familiar with or supportive of AI's use. The guiding principles for submitting documents to the Court highlight the importance of caution and the verification of AI-created content, advocating for a "human in the loop" approach. This strategy ensures that all legal references and analyses, generated by AI, are cross-checked against well-recognized and reliable sources, thereby maintaining the accuracy and trustworthiness of legal documents. This concerted move towards regulating AI's application in legal document preparation by the Federal Court and other Canadian jurisdictions signifies a cautious yet forward-thinking engagement with AI technologies within the legal profession.

AI + Copyright

In the last year, both the Canadian and US copyright offices have examined the copyright issues arising in the context of generative AI.  In Canada's case, the office published a "Consultation on Copyright in the Age of Generative Artificial Intelligence", discussing authorship and infringement issues.  The Copyright Act does not define the term "author", but Canadian copyright law suggests that a human must be involved in authorship for copyright purposes.

However, at least one Canadian copyright registration (an artistic work, entitled SURYAST) lists the "RAGHAV Artificial Intelligence Painting App" as an author, along with a human author, Ankit Sahni, who also happens to be an IP lawyer.  This creates an interesting precedent and has been described as a formal recognition in Canada of copyright authorship by a non-human technology. It's important to remember that there is very little in the way of examination or assessment of applications for copyright registration in Canada, so it's relatively easy to list any type of AI as an "author" in the application.  That's why the SURYAST registration shouldn’t be taken as a formal endorsement of any particular position by the Canadian Intellectual Property Office. It will be comparatively difficult to convince a court of the legitimacy of this registration, if it's ever tested.

In the US, the same application for SURYAST was rejected by the Copyright Office in December 2023, reflecting the ambivalence towards recognizing software as an "author".  The US Copyright Office also launched their own consultation initiative to examine American copyright law and policy issues in the AI context, and more formal guidance is expected in 2024. Historically, the US position has been clear that the Copyright Act protects original works of authorship, and that for these purposes "a work must be created by a human being."

Patent Law

The United States Patent and Trademark Office recently published guidance addressing AI's implications for patent inventorship. This guidance maintains that individuals named as inventors in a patent application are presumed to be the actual inventors, irrespective of any contributions made by AI that might otherwise qualify as inventorship. The crucial factor for inventorship determination is whether a human's contribution to the invention is substantial enough to merit inclusion as an inventor in the application. This policy, reflecting a balanced view shared by Canadian Courts, acknowledges both the complexities and the indispensable role AI plays in the current and future landscape of legal and intellectual property practice.

Embracing AI with Vigilance

The reactive measures in both Canada and the US underscore the legal profession's cautious yet forward-looking engagement with AI technologies. By setting clear guidelines for AI use, the legal profession seeks to harness AI's potential to streamline legal processes and improve access to justice without compromising the legal system's integrity.

This approach acknowledges AI's transformative impact on the legal profession, from automating mundane tasks to facilitating complex legal research. However, it also recognizes the inherent challenges, such as ensuring accuracy, preventing bias, and maintaining judicial independence.