STAT quoted Dr. Jay Anders in response to their questions to readers: What story do you think went underdiscussed in 2024? What should we all be paying attention to in 2025?
2024 lacked an honest conversation about accountability for AI errors in health care, as vendors rushed to deploy immature models without transparency around their reliability or limitations. While HTI-2 may provide some guardrails for generative AI in 2025, we’re witnessing companies recklessly positioning AI chat bots between patients and clinicians without addressing fundamental questions about medical liability when these black box algorithms inevitably make mistakes. The industry’s obsession with AI’s potential has overshadowed critical discussions about who bears responsibility when AI systems —
which can inexplicably change their outputs from one day to the next — contribute to patient harm.
