Artificial intelligence is already embedded across research tools, document workflows and knowledge platforms. The question legal teams are now facing is not whether to adopt AI, but how to do so in a way that builds confidence, manages risk and genuinely improves the work of legal professionals.
For Gilbert + Tobin, one of Australia’s leading law firms, this new way of getting work done is well established. The firm has been working with generative AI for several years, and according to Gareth Cantin, Head of Knowledge and Practice Innovation, the focus has moved well beyond experimentation.
“The question is less about adoption in isolation, and more about how we drive meaningful transformation across the firm,” Gareth says. “Adoption alone is not sufficient, value is only realised when it is sustained, intentional and aligned to how lawyers actually work.”
This mindset offers a lens for understanding what ‘trusted AI’ looks like in legal research.
Confidence starts with the lawyer, not the tool
For legal research in particular, trust does not come from automation. Gareth is clear: AI outputs must support, not supplant, legal judgment.
“Confidence in AI tools starts with recognising that the lawyer remains the critical decision-maker. The output is only part of the process; it’s how that output is interrogated and applied that really matters.”
This perspective reflects a broader shift across the profession. Lawyers are increasingly comfortable using AI to accelerate research and analysis, but only when they can see how an answer was constructed and which authorities it relies on.
That transparency is non‑negotiable – as Gareth puts it, reviewing the answer is almost as important as the answer itself.
“Any legal research tool needs to demonstrate it works while relying on authorities correctly in a transparent way, pinpoint citations and offer the ability to view referenced content simultaneously.”
In other words, trust is built when AI behaves less like a black box and more like a junior researcher whose work can be reviewed, tested and validated.
Why transparency matters in AI-driven research
After seeing a demonstration of Westlaw Advantage Australia, one capability stood out to Gareth as particularly aligned to how lawyers actually work: Deep Research.
What resonated was not simply speed, but visibility into the research process itself.
“Being able to follow the chain of process allows a lawyer to determine it’s on the right track,” he explains. “What’s valuable is that the output includes that detail as research steps, enabling reviewers to understand how the output is assembled.”
This approach mirrors the way legal teams already collaborate: one lawyer conducts research, another reviews it, and both need confidence in the authorities relied upon. Features such as KeyCite signals, AI‑generated summaries of underlying resources, and structured research trails help support a common understanding and establish a shared basis of fact.
For legal teams managing risk and accountability, this kind of auditable research path is critical. It allows AI to accelerate work without compromising rigour – a defining requirement for using trusted AI in law.
Accelerating work without replacing judgment
The most immediate impact of AI‑powered research, Gareth suggests, is straightforward: efficiency.
“It allows an acceleration of legal work which is underpinned by legal research. Lawyers can traverse content faster and in a way which provides structured and auditable research.”
But the efficiency gain does not come at the expense of expertise. At Gilbert + Tobin, AI is considered to be amplifying the firm’s capabilities rather than eroding them.
“The role of the lawyer remains central, interrogating outputs, applying judgment and ensuring the advice is sound.”
This distinction matters, particularly as firms weigh the risks of over‑reliance on generative AI. Trusted platforms must be grounded in authoritative content and designed to keep human judgment firmly in control.
Responsible adoption: guardrails, integration and visibility
Looking ahead, Gareth sees responsible AI adoption as a balance of opportunity and discipline.
Law firms are fundamentally knowledge businesses, and that creates both potential and risk. Confidentiality, data security and accuracy all demand careful governance.
“AI presents significant opportunity, but also real risk. Governance, guardrails and clear policies are foundational.”
Equally important is recognising that generative AI is just one part of the broader technology ecosystem. The goal is not to use AI everywhere, but to use the right tools for the right tasks at the right time – integrating AI research, workflow platforms and human expertise seamlessly.
“Successful adoption,” Gareth notes, “will depend on deeper integration between external systems like legal research platforms and firms’ internal resources – delivering trusted content directly into the environments where lawyers already work.”
At the same time, firms need visibility.
“We’re focused on building deeper visibility of how these tools are being used, so we can refine both the content and the services we deliver.”
Trusted AI is practical, not hype
The conversation around AI in law is often dominated by bold claims and sweeping predictions. What emerges from Gilbert + Tobin’s experience is something more grounded, and more useful.
Trusted AI today is transparent, auditable and grounded in authoritative content. It accelerates research without bypassing judgment. And it earns confidence not through novelty, but through reliability.
As platforms like Westlaw Advantage Australia continue to evolve, combining Deep Research capabilities with the trusted foundation of Westlaw content, the opportunity for legal teams is clear: smarter research, stronger confidence, and AI that works the way lawyers already do.
