How AI Misuse in Junior Lawyer Supervision Exposes Broader Risks
Recent court cases involving generative AI errors have exposed gaps in how law firms supervise junior lawyers and paralegals. The cases reveal that AI tools can mask poor work practices and create false confidence in less experienced staff, complicating the already difficult task of remote supervision.
Supervised legal practice typically lasts 18 to 24 months after admission. Despite its longevity, the arrangement remains a source of dissatisfaction for some junior lawyers. Most supervision issues never reach courts or disciplinary tribunals. The AI cases are useful precisely because they illustrate ordinary supervision failures rather than extreme circumstances.
How AI Errors Exposed Supervision Gaps
In Murray on behalf of the Wamba Wemba Native Title Claim Group v State of Victoria [2025] FCA 731, a junior solicitor working remotely used generative AI to generate citations for court documents. The solicitor deposed to preparing footnotes without access to either physical or electronic copies of the documents being cited. No one checked the work before filing.
The supervising solicitor described work as "performed collaboratively between team members." But collaboration without verification created a blind spot. The firm had no policy, guidance, or accepted practice for how remote work should be conducted or how checking should occur.
In the English case Ayinde v London Borough of Haringey; Al-Haroun v Qatar National Bank [2025] EWHC 1383 (Admin), a pupil barrister claimed to have received no supervision of her written work in the second half of pupillage. The High Court questioned whether those responsible for her training, supervision, and allocation of work had failed in their duties.
The Australian family law case Mertz & Mertz (No 3) [2025] FedCFamC1A 222 involved a paralegal, solicitor, and two counsel. The paralegal used AI to alphabetise references, then inserted some into footnotes-generating hallucinated material. The solicitor's explanation did not identify which AI program was used, what training the paralegal had received, or what guidance existed around AI use. The court ordered the solicitor to pay $10,000 towards the respondent's wasted costs.
The Trust Problem
AI complicates the trust relationship between supervisor and supervisee. Fluent writing and fast research completion can mask lack of skills or adherence to proper work methods. Someone using AI without disclosure may appear more capable than they actually are.
In several AI misuse cases, errors passed unnoticed by multiple people: solicitors, counsel, and even opposing counsel. This suggests supervisors cannot rely on their instincts about a junior lawyer's capabilities when AI tools are in play.
New lawyers face conflicting pressure. They must be efficient and technologically capable, yet fear reputational damage from AI misuse. This tension can push junior staff to conceal AI use rather than disclose it to supervisors.
What Supervisors Should Do
Create a clear supervision plan that explicitly addresses AI use. The plan should specify when AI may be used, how outputs must be reviewed, and who bears responsibility for final work product. This cannot be informal or assumed, particularly in hybrid or remote environments.
Supervision plans and AI policies must not operate separately. They need to be harmonised so supervisors know when AI may have been used and can review work accordingly. Final responsibility for work product lies with supervisors under professional conduct rules.
Understand how your junior staff actually work. Interrogating a colleague about their work methods may feel awkward, but supervisors must understand the practices of those they supervise. Are case law citations coming from authorised databases or chatbots? Was a document summarised by AI? Did work get produced suspiciously quickly?
Alarm bells should sound when a response is unaccountably wrong or work appears too polished to have been done without AI assistance.
Foster an environment where junior lawyers disclose AI use. New lawyers often misunderstand how AI tools work and what mistakes they can make. They lack experience to verify AI outputs. As they learn to practice law and use new tools simultaneously, instruction must cover both.
Normalising transparent AI use creates an environment for ongoing internal education about its ethical and practical limits. Supervisors might use AI themselves in their own work or as part of supervision-for example, to provide first-round feedback.
Ensure junior staff have support when work is challenging. Pressure to be efficient, combined with fear of reputational damage from AI errors, may produce uncertainty and a tendency toward concealment. Legal organisations must foster an environment where lawyers can discuss challenges in their work, especially those arising from time pressures.
The Broader Picture
Power imbalance inevitably exists between supervisor and supervisee. Supervisors cannot prevent junior staff from making mistakes or ethical lapses. But any supervisor with carriage of a matter has responsibility to oversee that work. Shifting blame to a more junior person-especially someone not yet admitted as a practising lawyer-is unfair and can damage that person's career before it starts.
Supervision matters more now, not less. AI makes it increasingly important to develop junior lawyers with strong technical skills, critical thinking about new technologies, and awareness of ethical obligations. The goal is to develop lawyers who are confident in their abilities and accountable for their work.
Learn more about AI for Legal professionals and consider the AI Learning Path for Paralegals to understand how junior staff should be trained on AI tools.
Your membership also unlocks:
