The Justice Department is currently investigating the healthcare industry’s use of AI in patient records that prompts doctors to recommend treatments. Subpoenas have been issued to pharmaceuticals and digital health companies to gain further insight into the role of generative technology in potential anti-kickback and false claims violations. This investigation is occurring as electronic health record vendors integrate more advanced AI tools to match patients with specific drugs and devices.
At this time, it is unclear how advanced the cases are and where they fit within the Biden administration’s initiative to promote innovation in healthcare AI while also regulating to ensure safeguards are in place. Some sources suggest that the DOJ is still formulating a strategy, as they are currently asking general questions in ongoing investigations.
The use of algorithms in AI is a key component of the technology, as it mines health data, identifies trends, and can match patients with certain conditions for potential remedies. While this has the potential to make healthcare delivery more efficient, it also raises concerns about AI abuse by companies seeking to profit. At least three publicly-traded pharmaceutical companies have disclosed receiving subpoenas from the DOJ related to electronic medical records, although no resolutions have been announced.
The investigation brings to light a 2020 criminal settlement with Purdue Pharma and its digital records contractor, Practice Fusion, for colluding to design alert systems that pushed doctors to prescribe addictive painkillers. This case serves as a cautionary tale for modern AI technology, with industry lawyers expressing concerns about potential legal liability for companies that benefit from the use of AI in healthcare.
This has led to discussion and speculation about where US prosecutors may be headed in their approach to AI, particularly considering the challenges associated with tracing criminal willfulness in automated AI systems. There is a cautionary approach to structuring AI in healthcare companies to avoid becoming targets of prosecution by DOJ and ensuring that the inputs and outputs are carefully managed.
Overall, there is a sense of cautious optimism about AI’s potential in improving healthcare, but also a recognition of the potential risks and challenges associated with its use in the industry.