In the world of legal AI, the assumption that all lawyers need the same level of guidance from technology is proving to be a fundamental flaw. This oversight, evident from recent empirical studies, suggests that legal AI tools are failing to cater to the diverse needs of lawyers at different stages of their careers.

During a series of classroom pilots conducted through Product Law Hub, involving an AI-based legal coach named Frankie, distinct preferences emerged between junior and senior legal professionals. The study, which combined quantitative engagement data with qualitative interviews, uncovered that while junior lawyers seek structure and clear guidance, their more experienced counterparts thrive on challenge and complexity.

The core of the issue lies in the uniform approach to AI design. Most legal AI systems are built with an implicit user model that suits early-career lawyers who desire structured prompts and checklists to navigate their cases. However, this model falls short for senior associates, counsel, or partners who prefer a system that stimulates critical thinking and presents complex scenarios.

The data from the pilot was unequivocal: junior users engaged more with the AI when it provided a clear pathway, whereas senior users disengaged when the system offered overly structured guidance, seeking instead a tool that could spar with them intellectually.

This discrepancy in AI utility is particularly problematic in law firms where the adoption of such technologies is often seen as mixed but acceptable from an external viewpoint. In reality, while junior lawyers may not be developing their judgment skills adequately, senior lawyers are likely not using the tool at all, finding it unsuitable for their advanced needs.

The insight from the pilot is clear: what is helpful in one stage of a lawyer’s career can be a hindrance in another. Early-career lawyers benefit from structured guidance, which should ideally decrease as they gain experience. On the other hand, seasoned lawyers need tools that introduce ambiguity and challenge their thinking, aspects vital for honing their judgment.

The study suggests that vendors and law firms need to rethink how they select and deploy legal AI systems. Instead of seeking a one-size-fits-all solution, there should be a push towards developing AI tools that adapt to the experience level and specific needs of the user. This approach, though more complex, would ensure that the technology is genuinely beneficial across all stages of legal careers.

As AI becomes increasingly embedded in legal practice, addressing this seniority issue becomes crucial. Ignoring the different needs based on experience levels does not only affect technology adoption but also impacts how legal talent develops critical professional skills.

Ultimately, the lesson here is unambiguous: legal AI doesn’t stumble because it lacks intelligence—it fails because it lacks differentiation. Firms must recognize that lawyers are not interchangeable in their needs and that a nuanced approach to legal AI development and deployment is essential for truly effective technology adoption in law practices.