The Federal Court of Australia is exploring the development of guidelines, or a practice note, to regulate the use of artificial intelligence (AI), particularly generative AI (Gen AI), in legal proceedings. This initiative, highlighted in a submission by the Law Council of Australia, aims to balance innovation with the risks posed by AI, such as the production of false or misleading material and breaches of confidentiality.
Gen AI tools, like ChatGPT, are increasingly used in legal research, drafting, eDiscovery, and evidence preparation. While these tools offer efficiency gains, they also present significant challenges, including “hallucinations” where AI generates inaccurate or fictitious legal references. The Law Council emphasises the need for clear boundaries, especially in evidence preparation, to ensure affidavits and expert reports reflect the witness’s own words and maintain integrity.
The submission suggests a hybrid approach, combining principles-based guidance with specific disclosure requirements. For example, practitioners might need to verify AI-generated content, while self-represented litigants could face stricter disclosure rules. The Law Council also highlights the risks of using AI for confidential information, noting potential breaches of privilege and confidentiality under existing legal obligations.
Courts in New South Wales, Victoria, and the UK have already introduced AI guidelines, offering models for the Federal Court to consider. The Law Council advocates for a proactive stance, recommending regular updates to keep pace with technological advancements.
As AI becomes ubiquitous in legal practice, the Federal Court’s guidance will play a critical role in safeguarding the justice system’s integrity while fostering responsible innovation. The Law Council has pledged ongoing collaboration to ensure the guidelines are practical and effective for all court users.
