Imagine a world where clinical trials are faster, cheaper, and more efficient. Sounds too good to be true? Well, it might not be, thanks to a groundbreaking development in artificial intelligence (AI). But here’s where it gets controversial: AI is now stepping into the role of adjudicating major adverse events in clinical trials, a task traditionally reserved for expert physicians. Could this be the future of medical research, or are we crossing a line? Let’s dive in.
On November 13, 2025, a study presented at the American Heart Association’s Scientific Sessions revealed that an AI model called Auto-MACE can evaluate critical events like cardiovascular death and stroke with accuracy comparable to human experts. Published simultaneously in the Journal of the American College of Cardiology (JACC), the research suggests AI could revolutionize clinical trials by streamlining processes and cutting costs. But this isn’t just about saving money—it’s about transforming how we conduct medical research.
And this is the part most people miss: AI doesn’t just replicate human judgment; it has the potential to improve upon it. According to Pablo M. Marti-Castellote, PhD, and his team at Brigham and Women’s Hospital, AI can reduce the volume of cases needing human review, slashing both costs and delays. By applying a consistent adjudication model across trials, AI could enhance reproducibility, a longstanding challenge in medical research. But how do we ensure this technology is ethical, transparent, and traceable? That’s the million-dollar question.
Alexandra Popma, MD, of the Cardiovascular Research Foundation, calls this study a ‘fantastic’ step forward but highlights the hurdles ahead. ‘The challenge is how we translate this into a product acceptable to regulatory agencies,’ she notes. The ethical implications are immense. How do we ensure AI meets standards for transparency and accountability? And what happens if it makes a mistake?
Auto-MACE was trained on data from five major cardiovascular trials, including INVESTED, DELIVER, and PARAGON-HF. In the PARADISE-MI trial, it confidently adjudicated 69% of deaths, 46% of potential heart attacks, and 81% of potential strokes. Its agreement with human clinical event committees (CECs) was impressive—97% for deaths, 89% for heart attacks, and 88% for strokes. Even more striking, both AI and human adjudication yielded similar results when evaluating the effectiveness of sacubitril/valsartan versus ramipril.
However, here’s where it gets tricky: Auto-MACE’s errors, though rare, were notable. For instance, it struggled with cases involving multiple cardiovascular issues combined with infections, like sepsis. It also misinterpreted previous heart attacks as new events and occasionally misclassified strokes based on prior brain imaging. These limitations raise questions: Can AI truly replace human judgment in such complex scenarios? Or should it serve as a complementary tool?
Looking ahead, the researchers suggest a hybrid approach, combining AI with human oversight. Early collaboration with regulatory agencies will be key to gaining acceptance for AI-generated data. Popma acknowledges the skepticism but remains optimistic. ‘AI addresses so many bottlenecks in clinical trials,’ she says. ‘It’s a fantastic challenge, and we need to find solutions.’
So, what do you think? Is AI the future of clinical trials, or are we moving too fast? Let’s spark a conversation. After all, the stakes couldn’t be higher—we’re talking about the health and lives of millions. Share your thoughts below!
