Big Picture: American Medical Association Calls for Federal Guardrails on Artificial Intelligence

Recent comments by the CEO of the American Medical Association (AMA) could be of interest for the disability community because of the comments’ implications about access to health care. The health field has been increasingly turning to artificial intelligence, such as algorithms, to inform its decision-making. This includes decisions on who should be prioritized to receive care. Because people with disabilities have lower access to health care compared to their peers without disabilities and are under-reported in datasets, this could be an interesting topic for readers who wish to connect disability issues and broader trends / discussions in health care.

As reported by Politico Pulse:

“AMA CEO: AI IS A MAJOR OPPORTUNITY…THAT NEEDS FDA, PHYSICIAN GUIDERAILS — Regulators and physicians must develop new frameworks and ethical codes to successfully deploy artificial intelligence, AMA CEO James Madara said in an interview on the sidelines of the HLTH conference in Las Vegas.

Madara pointed to research indicating that some AI programs showed racial bias on patient treatment advice — even when race was not an input in the algorithm. ‘These things need to have physician input,’ he told POLITICO’s Sarah Owermohle.

— That means it could be time for new guidelines from the AMA, Madara said, noting the organization’s first major code of medical ethics was developed more than a century ago. He also urged the FDA to take a leading role, acknowledging that ‘this is a big field and the FDA has limited capacity and is stretched pretty far already.’

— Data in general needs an overhaul — and doing so could help lower health costs , Madara concluded. “If I had one wish from the genie, it would be to make data better organized, clinically more meaningful, and liquid,” Madara said, arguing that 325 million Americans who use health services are effectively participating in clinical trials, even if they don’t know it.

‘It’s going on right now, we’re spending $3.5 trillion for it, and we’re just ripping the data up and throwing it away,’ he said.”