AM 640: ChatGPT in Divorce Court - Featuring Christopher Yu, Shulman & Partners

Artificial intelligence is increasingly shaping how people seek out legal information, a trend that Christopher Yu, Partner at Shulman & Partners LLP, recently discussed on AM 640's The John Oakley Show. The segment examined how tools like ChatGPT are influencing the expectations and decisions of individuals preparing for separation or divorce, often in ways that create confusion rather than clarity. As Chris explained, many clients are now arriving with AI-generated assumptions that do not align with Ontario family law, leading to misunderstandings, heightened stress, and unnecessary legal costs. The conversation underscored why AI, while useful for general information, cannot replace the nuanced, case-specific guidance provided by experienced counsel. It also highlighted the broader challenges lawyers face as AI becomes a more common part of the public’s decision-making process.

“People are doing a little bit of their own research first. Sometimes they say, ‘This is what ChatGPT told me about getting a divorce in Canada,’ and then they’re fixed to that idea. That isn’t always right.”
— Christopher Yu, Partner, Shulman & Partners LLP

During the segment, Christopher spoke about the growing trend of clients turning to AI platforms like ChatGPT for legal advice before, or even instead of, seeking guidance from a qualified lawyer. He explained that this shift is becoming increasingly common, with clients now presenting AI-generated suggestions as though they are authoritative legal strategies. While people often turn to ChatGPT to save money or gain a preliminary understanding of family law, Christopher noted that these tools frequently provide incomplete, misleading, or oversimplified information.

He compared the current moment to the early days of Wikipedia, where people were advised not to rely on it as a definitive source. He emphasized that while AI can offer surface-level explanations about divorce, child support, or general family law concepts, it cannot account for the nuanced and highly fact-specific nature of real cases. The risk arises when individuals rely on these outputs to challenge their lawyer’s advice or attempt to shape their case strategy based on AI rather than law.

One major consequence, Christopher noted, is increased legal costs. When clients insist on pursuing directions suggested by AI, their lawyers must spend time correcting misunderstandings, explaining why the information does not apply, or undoing the missteps caused by acting on incorrect guidance. He described situations where clients attempted to reduce child support or avoid dividing property based on AI-generated suggestions that were factually wrong or misinterpreted. In one example, a client insisted that incomes above $150,000 were exempt from child support guidelines because ChatGPT had suggested it—a notion Christopher had to carefully disprove.

He also discussed the psychological impact. Because AI is designed to mirror a user’s tone and assumptions, it can reinforce misconceptions rather than challenge them. He explained that users may unknowingly prompt AI tools in ways that produce the exact answers they want to hear, giving a false sense of confidence in incorrect information. This can heighten anxiety, erode trust in legal counsel, and lead people further away from solutions grounded in Ontario family law.

When asked about the future, Christopher acknowledged the rapid pace of AI development but expressed doubt that it could replace the expertise, judgment, and human understanding required in family law. The interview ultimately underscored the same message: AI may be helpful for general learning, but decisions about separation, support, and parenting require careful legal advice that accounts for the unique details of each family’s circumstances.

Listen to the full AM 640 Radio segment here.

This media appearance is part of Shulman & Partners LLP’s ongoing contributions to Canadian family law discussions. Explore more of our media features in our In the Media  archive.