The fax machine is not a joke in American healthcare. It is a workflow layer, and AI startups are learning that old plumbing can still hide a very modern venture opportunity.
The most frustrating bottleneck in U.S. healthcare may not be a shortage of clinicians or a lack of software. It may be the pile of documents still moving through fax lines between primary care offices, specialty clinics, payers, pharmacies, hospitals, and medical equipment suppliers.
That sounds absurd until you look at what the fax is actually doing. Referrals move through it. Prior authorization packets move through it. Medical records, discharge paperwork, lab orders, insurance cards, and handwritten notes all still pass through document rails that were never designed for searchable, structured, automated healthcare. The fax machine is not hanging around because doctors love old technology. It survives because it works across incompatible systems, satisfies compliance habits, and asks almost nothing from the sender.
As TechCrunch recently noted, that unglamorous reality is starting to look interesting to venture investors. The opportunity is not to laugh at faxing. It is to understand that a massive regulated market still depends on messy, unstructured administrative work, and that work is expensive. U.S. healthcare administrative costs are often estimated in the hundreds of billions of dollars a year, with some estimates running as high as $1 trillion. That is not a side issue. It is part of why patients wait, clinicians burn out, and providers lose revenue on tasks that should have been routine.
The new class of startups is not trying to persuade every clinic to rip out its systems. That is the important part. Healthcare has already spent years digitizing itself, yet much of that digitization has produced more portals, more logins, and more fragmented data. A specialist may use one EHR, a referring physician another, and a payer a separate authorization portal. Faxing remains the fallback because it crosses those boundaries without an integration project.
Coral is one of the clearest recent examples. The New York startup raised $12.5 million in a round led by Lightspeed and Z47 to automate administrative workflows for specialty providers. Its pitch is pragmatic: connect to existing EHRs, fax lines, and payer portals, then use AI to process patient intake, prior authorization, fax handling, billing, and follow-up communications. The company says its models can handle handwritten faxes, scanned insurance cards, prior authorization templates, and payer portal screens with 99.7% accuracy, while cutting complex patient intake to under five minutes.
That matters because specialty care is where the paperwork turns painful very quickly. Durable medical equipment suppliers, infusion centers, specialty pharmacies, and radiology groups do not just need a form filed neatly. They need the right document read, the missing field chased, the payer contacted, the patient scheduled, and the case moved before treatment is delayed. In that environment, an AI system that can read and route documents is not a nice dashboard. It can decide whether revenue is collected and whether a patient gets seen on time.
Tennr has been attacking a similar problem around referral workflows, with backers including Andreessen Horowitz, Foundation Capital, Y Combinator, and others. Insight Health, which raised an $11 million Series A led by Standard Capital with participation from Pear VC, Kindred Ventures, Eudemian, ElevenLabs, and 43, is also working around the administrative layer with AI assistants that handle referrals, patient history, and scheduling for routine procedures. These companies are different, but the pattern is the same: they are going after the work that sits between clinical intent and actual care.
The patch may still be valuable
There is an obvious criticism here. Automating fax workflows can look like putting better software on top of broken plumbing. If healthcare really needs interoperability, why spend venture dollars making the fax queue smarter?
The answer is that bridges can be good businesses. CMS has moved to standardize electronic claims attachments, with a final rule set to take effect on May 26, 2026, and compliance required within two years. The agency has estimated annual savings of about $781 million from modernizing that exchange. But healthcare rarely changes in one clean motion. Even when new standards arrive, clinics, payers, and vendors need years to adjust, and the messy middle is where startups can create value.
AI is well suited to that middle because the problem is not just transport. It is interpretation. A faxed referral may contain a diagnosis, insurance details, clinical history, missing signatures, and contradictory notes. Someone has to decide what it means and what should happen next. Large language models, document AI, and workflow automation are useful here because they can extract meaning from imperfect inputs, summarize the case, route it to the right queue, and trigger the next administrative step.
This is also why the category can become defensible. General-purpose automation tools struggle when every customer has different forms, payer rules, clinical specialties, and exception paths. Founders who embed deeply in one workflow can build data, integrations, and operational knowledge that are hard for a horizontal tool to copy. In healthcare, the boring details are often the moat.
The bigger question is whether these startups eventually help replace faxing or simply make it more tolerable. In the short run, the answer may not matter. Patients need referrals processed now. Providers need fewer denials now. Staff need relief from document queues now. The founders who understand that urgency can build useful companies before the grand interoperability future arrives.
The fax machine may eventually disappear from healthcare. Until then, the smarter bet is that the winners will be the companies that treat it not as a punchline, but as evidence of where the system still hurts.
Also read: Microsoft’s OpenAI anxiety shows cloud loyalty has limits • Meta is learning that AI transformation has a workplace cost • Chris Hohn cuts Microsoft as AI winners face a harder test
