AI in Medical Documentation: The Licensing Risks for Michigan Healthcare Providers

By: Jesse Adam Markos, Esq.
Wachler & Associates, P.C

AI is changing the way healthcare providers create medical records. AI-powered platforms can produce comprehensive clinical notes that summarize patient visits in a matter of seconds. These tools can significantly reduce administrative burdens and allow healthcare providers to spend more time with patients focusing on clinical care. While this technology offers tremendous promise for improving efficiency and reducing documentation burdens, Michigan healthcare providers should exercise caution and incorporate best practices before relying on these tools in their practice. As the provider, they remain legally and professionally responsible for every word contained in the medical record. And improper reliance on AI-generated documentation may expose providers to potential disciplinary action by the Michigan Department of Licensing and Regulatory Affairs (“LARA”).

Michigan’s Public Health Code provides the legal framework under which AI-generated documentation is evaluated. More specifically, pursuant to MCL 333.16221, LARA may investigate and discipline licensees for inadequate recordkeeping, negligence, incompetence, lack of good moral character, fraud or deceit in obtaining third-party reimbursement, and unprofessional conduct. And medical records frequently serve as the primary evidence in these investigations. Regardless of how a medical record is created, the licensed provider remains ultimately responsible for its accuracy and completeness.

AI-generated documentation presents an updated version of a documentation problem frequently encountered in licensing investigations: the overreliance on copying and pasting of prior medical records. The widespread adoption of electronic health records made it easy for providers to copy forward portions of prior clinical notes. For years investigators have identified situations where providers repeatedly copied prior notes without ensuring they remained accurate. As a result, outdated histories, physical examination findings, inaccurate medication lists, and conflicting treatment plans were inserted in the medical record. In some cases, these inaccuracies formed the basis for allegations of negligent recordkeeping, false documentation, or improper billing. And the convenience of technology did not excuse the errors.

AI-generated documentation has the potential to create even greater risks as it can generate new content. While AI can be impressive, it is also capable of making significant errors. For example, an AI-generated note might document physical examination findings that were never performed, insert diagnoses that were never discussed, misstate medication dosages, or create patient histories that differ from what was actually reported. It may also generate treatment recommendations that do not reflect the provider’s actual clinical judgment. If these inaccuracies become part of the permanent medical record, they can jeopardize patient safety, compromise the integrity of the medical record, and later be viewed as false or misleading documentation.

AI has the potential to improve efficiency, reduce burnout, and allow providers to devote more time to patient care. However, it should only be used with significant oversight and scrutiny. As LARA confronts the growing use of AI in clinical practice, providers must remember that they are ultimately responsible for the accuracy, completeness, and integrity of the medical record. And by maintaining appropriate documentation practices, exercising independent professional judgment, and utilizing only secure, HIPAA-compliant AI technologies, Michigan healthcare providers can significantly reduce the risk of any potential licensing issues. For additional information or assistance, please feel free to contact Wachler & Associates at (248) 544-0888.