“The only limit is your imagination” 

— Big Hero 6

In the film, the line isn’t about humans trying to outdo machines. It’s about humans learning what to do once the machine helps them go further. Baymax gives Hiro space to think, build, and solve problems that actually matter. 

That is exactly how AI is unfolding inside legal work.

Going by the general discourse, you would think the legal profession is standing on the edge of a dramatic cliff, as AI is either about to replace lawyers entirely or fail spectacularly trying. As we’re part of the journey to build the future of how legal works, this did not sit well with us. 

We set out to have real conversations with practitioners, and they sounded calmer. More practical. Sometimes, even oddly ordinary.

No one described waking up to a transformed career overnight. Instead, they described small moments: a contract understood faster than usual, a summary that removed an hour of reading, or even a second opinion rather than replace their thinking.

And the more they spoke, the clearer the pattern became. AI was not replacing legal expertise. It was rearranging where that expertise was spent.

The profession wasn’t shrinking. It was shifting.

What follows are five themes that keep resurfacing, shaping how legal is quietly evolving in practice rather than prediction.

Key Theme 1: The Work Changed Before the Role Did

The first theme appeared almost immediately, and it surfaced in nearly every conversation: the fear of replacement. 

No one described AI doing their job. They described AI clearing the space around their job.

Tasks that once took patience, be it combing through clauses, comparing versions, or extracting information from long documents, began to feel lighter. The time once spent locating information started turning into time spent interpreting it.

That distinction mattered more than anyone expected.

Legal expertise, they realised, had never really lived in the mechanical effort. It lived in deciding what mattered. The negotiation strategy. The judgment call. The risk tolerance is shaped by context and experience.

AI did not remove the thinking. It removed the search. And that became a consistent reframing: AI was not eliminating value. It was isolating it.

Kerry Sheehan from Doral Renewables LLC puts it very well when asked about the myths of legal in AI —

“... that AI can replace lawyers or legal workers. So in my experience, it replaces inefficiency, not expertise. And the best results come when AI augments good legal judgment instead of trying to substitute for it.”

Key Theme 2: Time Was Reassigned to Something Better

Everyone spoke of how AI freed up their time, but that didn’t mean lawyers were idle in the time that was freed up. When repetitive effort began shrinking, the team became early.

Early in conversations. Early in planning. Early in decisions, they were once pulled into only after problems appeared. They became proactive, from reactive. 

Natalie Salunke puts it beautifully when asked about how her time has freed up using AI —

“So I think it's more about, yes, getting the time back to invest in sort of strategy per se. So the stuff that you always want to do, you have on your list, it's project based…Whereas before you know that you're gonna have to chip away with it. It could take months, you know, even, you know, half a year, say. But actually knowing that you've got these tools there, I think it makes progress feel a bit more, um, satisfactory…You know what? We can see much quicker the fruits of our labor. And I think that helps propel people to feel like, you know what? Let's just keep going. We can do this. We can see the results quickly. Um, and it's less daunting.” 

From the conversations, it was easy to interpret that legal is moving from interpreter to architect.

Over time, this changed how others interacted with legal teams as well. Business stakeholders began engaging earlier because responses were faster and clearer. Conversations became collaborative instead of corrective.

AI did not just accelerate tasks. It expanded participation.

Key Theme 3: Trust Didn’t Disappear, It Became Intentional

Despite the optimism, one thing remained constant across every discussion: no one handed responsibility to AI. Not once. 

Verification stayed embedded in the workflow, but it stopped feeling like redundant labour and started feeling like quality control over a strong first draft rather than creation from scratch. They learned where AI performed reliably and where human interpretation was essential. 

Pattern recognition and structured summaries? AI is Helpful. Contextual nuance, negotiation tone, or risk balancing? Always a human territory.

Rathini Stear compared it to working with a junior colleague — helpful, capable, occasionally wrong, and always requiring review before anything left the room.

It was great to see that the responsibility never moved; only the starting point did. Instead of staring at blank pages, lawyers now begin with material to refine. The human-in-the-loop didn’t weaken, there was much more clarity now that before. And best of all, AI did not change accountability, it clarified ownership.

Alessandra Colaci says —

“Yeah, I think when people think to themselves, is this something that my judgment as a lawyer can be still utilized. And that's the thing that even gets them excited that's what they're trying to do…that should still be done by a human or by me because it's, you know, again, AI needs a little fewer checks and balances. And so it's not to the point where we can just trust it 100%.” 

Key Theme 4: The Future Wasn’t One Tool, It Was Many Small Ones

From our conversations, one thing was clear - there is not perfect tool to navigate the legal profession. When asked about what their AI legal assistant in 2026 would look like, everyone was quick to respond with their expectations. No one described finding a single perfect AI platform that solved everything. In fact, the opposite kept appearing: usefulness came from combining tools, each good at something specific.

One for drafting. Another for summarising. Another for note-taking. Another for brainstorming. Another for structuring knowledge.

Even though at the beginning it would look like only a few minutes saved, the result was that it actually saved the team hours on end by the end of the week. 

The professionals who benefited most were the most curious. They experimented, adjusted, replaced tools when needed, and gradually built personal workflows that matched how they thought.

Across the board, leaders used AI less like software and more like infrastructure; something woven into daily habits rather than accessed only for special occasions. What mattered was not finding the smartest tool. It was understanding the task clearly enough to choose the right assistance.

They stopped asking when AI would be ready. They started asking what small part of today could be easier.

Key Theme 5: Legal Work Became Less Solitary 

Traditionally, legal teams were invited into conversations at the moment decisions were almost finished, the contract drafted, the deal shaped, the messaging agreed, and their role was to review, adjust, and protect. The structure of the work unintentionally placed lawyers at the perimeter of the business, interacting with teams in short, high-pressure exchanges rather than continuous collaboration. Legal operated carefully, but separately.

This ties back into all the time that was saved in mundane tasks; legal now had time to be active participants in the business. 

It allowed lawyers to participate earlier in the planning itself. Several professionals described being pulled into conversations before commitments were made, not to approve, but to help shape structure, rollout, and expectations across teams.

Lawyers were now helping determine direction. 

AI also helped them explore those directions beforehand, outlining scenarios, comparing approaches, and clarifying implications; so discussions became joint problem-solving rather than defensive checkpoints. Legal advice felt less like a stop sign and more like navigation.

The function didn’t expand by mandate. It expanded in capacity.

With operational pressure reduced, legal moved naturally into cross-functional strategy, becoming part of how decisions were formed rather than where they were validated, and the work became less about approval and more about shared planning.

‍

Judgment will always stay human

Across these conversations with women leading the way, the transformation never sounded dramatic while it was happening. No one described a single moment where everything changed. Instead, they described noticing, weeks later, that certain frustrations no longer appeared. That emails were clearer. That reviews were calmer. That discussions started earlier.

AI did not replace lawyers, nor did it simply make them faster typists. It shifted attention toward thinking, advising, structuring, and communicating, the parts of legal work that had always carried the most value but often received the least uninterrupted time.

And perhaps the most important insight was this: the lawyers who benefited most were not the ones trying to hand over responsibility, nor the ones refusing to engage. They were the ones treating AI as a collaborator in preparation but never in judgment.

The work changed first.

Then the role expanded.

Legal is not just being automated.

It is becoming augmented and more visibly human at its center than before.

‍