For a growing number of law firms, AI-enabled legal research tools have moved beyond the experimental stage

In legal practice, speed matters but certainty is critical. This tension sits at the heart of how many law firms are approaching generative AI. Attention is turning to how it can be used to speed up legal research while maintaining trust, the ability to verify its output and professional standards.

At Holding Redlich, a commercial law firm, the chief knowledge officer Keren Smith says the business is taking a strategic approach, selectively adopting AI tools into legal research processes, guided by healthy skepticism.

“We haven’t rushed into it by any means,” she says. “We put our toe in the water early, prior to fully integrating AI into our legal research workflows.”

Trust comes first
That measured approach reflects a broader reality for law firms. Lawyers want to deliver faster answers and advice to clients, while also performing strongly in court, and being able to confidently account for every authority cited and conclusion reached.

Smith says Holding Redlich is very particular about the AI tools it adopts: they must be verifiable platforms, such as those offered by Thomson Reuters, grounded in trusted research. The ideal tool facilitates fact-checking to help ensure hallucination-free content.

Before rolling out new tools, the firm developed an internal AI usage policy that staff must agree to before gaining access. Risk management, Smith says, has always been a priority. “Prior to rolling out any AI tool, you need to have a risk management strategy in place.”

Holding Redlich gives lawyers access to AI-enabled legal research tools as well as ongoing training so they can effectively use them. Lawyers are taught traditional research methods alongside AI-enabled workflows, including how to frame prompts effectively and assess results critically.

“We also train lawyers in precision prompting, so that they know which terms and phrases will return relevant research results. Knowledge of correct legal terminology is essential.”

Smith says Thomson Reuters’ tools have helped to build lawyers’ confidence, and have particularly helped get junior lawyers up to speed quickly. Content on the legal research platform Westlaw Advantage Australia is updated daily and is linked to verifiable source material, ensuring research is current and accurate.

“Thomson Reuters has always been a reputable provider of legal content whose services we have subscribed to and trusted for many years.”

Faster, verified research
Once those foundations are in place, the impact of AI can go beyond improving efficiency; it can reshape how legal research is conducted day to day.

Smith points to a recent example involving a newly arrived lawyer who tested Westlaw Advantage Australia on a live task. Research that would ordinarily have taken more than three hours took about 10 minutes. In busy practice, time savings such as that can quickly add up.

Just as important, Smith says, is the ability to verify outputs through linked citations and authoritative source material, rather than relying on opaque responses.

James Jarvis, the vice-president of Westlaw product management at Thomson Reuters, says this is where many AI conversations miss the point.

Headshot of Vice-president of Westlaw product management James Jarvis, Thomson Reuters
Vice-president of Westlaw product management James Jarvis, Thomson Reuters

“When people talk about AI, they’ll focus on the fact that you can do things faster or easier with AI, but as professionals, lawyers need to be confident that the information and material that the AI is responding with is accurate,” Jarvis says.

The real value is not the research report that’s generated, but how quickly it guides lawyers to the underlying authorities they must ultimately rely on, Jarvis says. AI may accelerate the path, but it does not replace the need for primary sources. He says the benefit of using advanced reasoning models and agentic AI – within a closed system such as Westlaw Advantage Australia – is that the technology is trained to respond using only the Australian legal information within the platform, eliminating the possibility of unreliable results.

“Critically, the point of this exercise is to get the lawyers to the relevant primary information more effectively. The goal isn’t to stop with the research report; it’s to use that as the springboard to the relevant primary law that they have to cite in court.”

In practice, research outputs are linked back to the source material. Users can see AI-generated summaries explaining why a case or statute is relevant, view the specific passages relied on, then move directly to the court-ready source document. The platform also flags where cases have later been considered or treated by the courts, helping lawyers assess whether the precedent remains sound.

Verification must be designed into the workflow, Jarvis says, because even with the best intentions, checks can be missed or pushed aside when time runs short.

“With Westlaw Advantage Australia, verification isn’t an extra step; it’s baked in,” Jarvis says. “Tools like our Litigation Document Analyser enable lawyers to check citations in real time, flagging references that can’t be matched to authoritative sources.”

The competitive edge
For Smith, the benefits of AI go well beyond faster research. By reducing time spent on routine tasks, the tools help lawyers do more work efficiently, while freeing them to focus on higher-value priorities such as client service, business development and strengthening relationships.

In a competitive market these qualities can differentiate one firm from another.

Despite the profession’s caution around AI, Smith says Holding Redlich’s lawyers have been eager to embrace tools such as Westlaw Advantage Australia. After 26 years in law, she sees AI’s growing role in legal practice not as a leap into uncertainty, but as genuine progress – provided it is supported by strong governance, trusted content and rigorous training.

For firms still deciding whether to move forward with AI tools, she says waiting too long can carry risks of its own.

“There’s certainly a danger in not adopting it,” Smith says. “The fact of the matter is, it’s here and the real risk lies in not using it, rather than using it.”

• This article was originally published in THE GUARDIAN (UK) and was paid for by THOMSON REUTERS

(C) CITY LAWYER Magazine. All rights reserved. To join our Channel, click here. Click here to join our WhatsApp chatroom. Contact us at citylawyermag@gmail.com or 081-3838-0083.