Conference Insights by Peter Saunders, Maree Beare MMPP, and Chelsea Gordon

At the ARCS Annual Conference, the session “AI in Action: Bridging Innovation and Practicality in Clinical Trials” brought together diverse perspectives on one of the most pressing topics in the life sciences sector. Moderated by Peter Saunders, Sector Consultant, the panel featured Maree Beare, CEO of Clinials ; Chelsea Gordon, Legal Lead-AI advisor at Minter Ellison; and Saurabh Jai, AI innovator and data strategist. Together, they explored how artificial intelligence is not only reshaping clinical trial design and delivery, but also challenging how the sector views ethics, inclusivity, governance, and value.

The discussion began with a shared observation: artificial intelligence is no longer theoretical in clinical trials. It is already in action. “We’ve moved beyond generic models like ChatGPT,” said Jai. “Now we’re working with clinical trial-specific models trained on decades of global trial data. These models can identify the optimal design features, inclusion criteria, and even pinpoint the most promising trial sites.” This specificity, he argued, is where AI demonstrates real, measurable impact.

Beare offered a practical example. Her company Clinials      uses AI to extract data from protocols and other trial documents to generate operational and patient-friendly content such as summaries, landing pages and pre-screening materials. “We’ve seen reductions of up to 70 percent in document review times,” she said. “That translates not just into operational efficiency, but improved communication across sponsors, CROs and sites. Ultimately, it helps support more inclusive recruitment.”

Gordon, speaking from a legal and regulatory perspective, stressed the importance of distinguishing between traditional machine learning and newer generative AI. “The real shift, and where the risks start to emerge, is with AI that can create new content,” she said. “That raises questions around accuracy, accountability, and bias.” Governance, she explained, must be seen as foundational rather than optional. “If you're using AI in a high-risk setting like clinical trials, AI governance is your ticket to participate.”

The panel agreed that the adoption of AI must be intentional. “It’s easy to get caught up in the hype,” said Beare. “But you need champions in your organisation and a clear understanding of what problem you’re solving.” Whether it's addressing site burnout, recruitment diversity or trial design complexity, panellists argued that AI is a tool to augment rather than replace human expertise.

Jai offered another vivid example of this augmentation. “There’s so much value in trial site selection. AI can rank the best-performing sites globally for a specific indication based on past trial data. These aren’t assumptions. This is what the data is telling us. And it’s already being used by sponsors to de-risk trials from the start.”

Gordon reiterated the need for caution, particularly around patient data. “Australia already has a complex web of laws including privacy, consumer, intellectual property and therapeutic goods legislation. Most people don’t realise they already apply to AI,” she said. “Just because you can, doesn’t mean you should. Think of governance as people, policies, and processes. Define your approach up front and build guardrails that enable safe innovation.”

As the conversation turned to the future, there was optimism about AI’s trajectory. All panellists agreed the next two to five years would see rapid evolution, particularly in the rise of AI agents that mimic specific trial roles, more refined predictive analytics, and potentially regulatory expectations for AI-based trial simulations before involving human participants.

But the strongest message was one of balance. “AI is not a silver bullet,” Saunders concluded. “But when applied with intention, inclusivity, and responsibility, it is a powerful tool. The challenge for our sector is to keep learning, testing, and sharing, because that is where the real progress happens.”