Powered by
MOMENTUM
MEDIA
As legal teams race to move beyond experimentation and implement AI at scale under mounting pressure to keep pace, many are making critical missteps that could leave firms exposed to significant long-term consequences in an increasingly AI-driven legal landscape.
As artificial intelligence becomes increasingly embedded in the legal profession, legal teams are under mounting pressure not just to experiment, but to embed AI into meaningful, day-to-day practice at scale.
Yet in the urgency to move fast, many are making critical missteps that risk undermining the very long-term value and competitive advantage AI is meant to deliver.
Speaking on a panel at the Lawyers Weekly Corporate Counsel Summit, Andrew De Celis, executive general counsel at Telstra, and Clementine Fox, general manager at Luminance, unpacked the most common pitfalls in AI adoption and outlined what legal teams must do to avoid them and get implementation right.
Stop treating AI as a rollout
One of the biggest mistakes De Celis highlighted in legal teams’ approach to scaling AI is the tendency to treat it as a routine technology implementation, rather than recognising it for what it truly is – a fundamental shift in how legal services are delivered, structured, and transformed.
“Are you treating AI as a technology rollout? This will tell you a lot about your mindset,” De Celis said.
“If you’re treating AI as a technology rollout, you’re missing the scale of the opportunity.”
Rather than focusing solely on implementation, De Celis argued that legal teams should see AI as more than a tool for efficiency, recognising it instead as a capability that expands human potential and fundamentally reshapes how legal professionals create and deliver value.
“AI lifts the ceiling for human potential. To unlock that potential requires more than flawless tech implementation,” De Celis said.
“It requires investment in new capabilities and reimagining how we work and deliver value.”
Beyond isolated AI tools
For Fox, one of the most common failures in scaling AI is legal teams treating the technology as a patchwork of disconnected tools rather than embedding it into the fabric of their day-to-day workflows.
“One of the most common mistakes organisations make when trying to scale AI is treating it as a collection of isolated tools rather than embedding intelligence into the way teams already work,” Fox said.
“Many businesses experiment with AI at individual stages of a process, but as work moves between teams and systems, valuable context is lost.”
Given that precedent and institutional knowledge are the lifeblood of legal teams, Fox warned that this fragmented approach can quickly undermine AI’s effectiveness, stripping systems of the continuity and organisational memory needed to deliver reliable, trusted insights.
“In legal teams especially, scaling AI successfully depends on more than simply deploying large language models. Teams need AI that can reason with context, retain organisational memory, and deliver transparent outputs professionals can trust,” Fox said.
Rather than chasing the latest AI tools, Fox argued that the organisations seeing the greatest success are taking a far more deliberate approach, embedding AI into existing workflows, focusing on clearly defined use cases, and building governance and trust from the outset rather than as an afterthought.
“The organisations seeing the strongest results are taking a practical, targeted approach by starting with clearly defined use cases and embedding AI into existing workflows rather than introducing disconnected tools,” Fox said.
“Successful teams are also prioritising governance, transparency, and human oversight from the outset so professionals can trust and validate AI outputs with confidence.”
Focusing on the ‘winning tool’ instead of capability
Another major trap De Celis identified is the growing belief that successfully scaling AI is simply a matter of backing the ‘right’ or ‘winning’ platform.
He argued that with new AI platforms and legal technology products emerging almost weekly, many organisations are wasting valuable time trying to pick a future winner rather than focusing on how to extract value from the technology available today.
“The winning technology won’t determine the winning team. Models will improve, vendors will change, and today’s ‘it tool’ will quickly become tomorrow’s baseline,” De Celis said.
Instead, De Celis argued that long-term success will not come from the tool itself, but from how effectively legal teams build systems, processes, and skills that allow them to harness AI regardless of the platform being used.
“The legal teams who get the competitive edge will be those most effective at harnessing AI (no matter what tools they use),” De Celis said.
“AI tools will get leapfrogged. The system you build around it endures.”
