Insights
Faxes in the Age of AI
A few weeks ago, I attended a conference session titled “Implementing FIHR While Faxing Is Still a Reality in Prior Authorization.” It was terrific — with sophisticated techniques, a fax can be essentially reverse-engineered into a much more useful electronic document — but also gave me a moment of pause. How is it that we are in a world with such extremes of technology “haves” and “have-nots” that some in the health care sector — particularly nursing homes, mental health providers and small or rural clinics — are still filling out paper forms and feeding sheets into fax machines, while others have cutting-edge technology to parse the results into something they can reasonably use?
By Emily R. Transue,
June 25, 2025
Health IT
A few weeks ago, I attended a conference session titled “Implementing FIHR While Faxing Is Still a Reality in Prior Authorization.” It was terrific — with sophisticated techniques, a fax can be essentially reverse-engineered into a much more useful electronic document — but also gave me a moment of pause. How is it that we are in a world with such extremes of technology “haves” and “have-nots” that some in the health care sector — particularly nursing homes, mental health providers and small or rural clinics — are still filling out paper forms and feeding sheets into fax machines, while others have cutting-edge technology to parse the results into something they can reasonably use? Beyond faxes in particular, what other gaps mark this divide? Are we on a course to this getting better, or worse?
For those who are digitally sophisticated — i.e., typically those with financial resources, technical expertise, and the bandwidth to figure out and implement new systems and workflows — artificial intelligence and other new technologies are rapidly becoming transformative. Electronic scribes and ambient listening can reduce provider administrative burden and alleviate burnout. Relevant patient information can (or soon will) be extracted from health records (even old scanned or faxed documents!) and presented to clinicians in ways that support improved diagnostic decisions, evidence-based treatment plans, and identification of care gaps, risks and unmet needs. There are risks in all of these advances: overreliance on imperfect AI systems, data security breaches, AI-generated “hallucinations,” AI perpetuation of systemic bias. Managing these risks requires more resources and sophistication, which further differentiates the haves from have-nots.
Meantime, others are still sending faxes. Some are still using paper records or struggle to have a functional EMR, much less AI. Staffing crises and financial constraints at many nursing homes and behavioral health agencies are so severe that merely keeping the “doors open and lights on” is a daily struggle. If AI applications are adopted by hospital systems and others with more resources and thus improve workloads and administrative complexity for their staff, those already struggling with workforce — skilled nursing facilities, behavioral health, rural providers — will face even more difficulty competing for workforce.
As the health care system increasingly incorporates powerful new AI tools, care will need to be taken to ensure that less-resourced components of the system have access to and support in implementing these tools. Otherwise, inequities may be increased and the most fragile parts of the health care ecosystem will be even more at risk. It’s wonderful that we can use AI to interpret faxes. But it would be much better if we didn’t need to.
Author / Contributor(s)
Recent Insights