AI in healthcare - getting it right, and stopping it going wrong

This week we’re focusing on how healthcare is using AI systems. The NHS is firmly set on a course of AI integration, with NHS England committing £10 billion of taxpayer funds to the cause over the next three years. The pressure to implement AI in our hospitals is intense. But, how can healthcare leaders know what makes for a good AI adoption strategy? Victor Montori and Jon Tilburt are physicians from the US’ Mayo Clinic. They join us to lay out what they are calling the ‘five red flags’ that any organisation needs to consider in its AI rollout.   The UK is already having this conversation. The National Commission into the Regulation of AI in Healthcare released its initial report in June of this year. Among its findings was a call for a ‘lifecycle-based approach’ to regulation. To explain what this means, as well as how doctors can be equipped to ‘surveil’ AI systems as conditions evolve, we’re joined by the co-authors of a new BMJ editorial - Mark Ball from the University of Derby, and Elvira Perez Vallejos from the University of Nottingham.   And lastly, we talk about a research paper from last week that sparked off some concern. Researchers at Stanford University in California have used generative AI to design novel bacteriophage viruses. But, what have these scientists actually accomplished, and how can we put the right safeguards around the technology? Simon Clarke, associate professor in cellular microbiology at the University of Reading, explains the breakthrough and lays out its real implications.   References: We can, but should we? Red flags for hasty, low integrity AI integration into healthcare Regulating AI in healthcare: a moving target Developments in AI designed viruses demonstrate why such research must be done in the open