The phrase “drug repurposing” sounds tidy until you watch how rarely the world actually funds it.
Personally, I think we keep talking about rare diseases as if they’re a scientific problem, when a huge part of the bottleneck is political, financial, and operational. Every Cure’s approach—especially its “disease-agnostic” and AI-assisted model—forces an uncomfortable question: what if we’re treating rare disease like a series of one-off petitions, instead of building a scalable system that simply looks?
What makes this particularly fascinating is that the organization isn’t trying to be a charming matchmaking service between a single disease community and a single drug hypothesis. It’s trying to restructure the pipeline itself, from idea to patient, and that kind of systems thinking is exactly what rare disease medicine has lacked.
Rare diseases aren’t rare in reality
Start with the basic math: there are over 10,000 rare diseases, and most don’t have a cure. That fact alone should make anyone uneasy. In my opinion, the most misleading thing we do in public conversations is imply that the scarcity of patients is a scarcity of importance. It isn’t.
What many people don’t realize is that “rare” often means “economically ignored,” not “biologically uninteresting.” When patient populations are tiny—sometimes dozens—traditional drug development becomes financially irrational. So the system leans on the assumption that if there isn’t profit, there shouldn’t be effort.
From my perspective, that assumption is morally flawed and also strategically dumb. We’ve already seen countless medical breakthroughs arrive from unexpected angles; the problem is that the current incentive structure mostly blocks those angles in rare disease.
And if you take a step back and think about it, repurposing is the logical counterpunch: we don’t need to start from nothing every time. We can start from biology we already understand and manufacturing pathways we already have.
The “two ways” of repurposing—and the one everyone avoids
Every Cure describes two possible routes: either invite disease groups to request help, or use AI to identify the most promising drug–disease matches across the entire universe of drugs and diseases.
I personally think most organizations choose the first option because it feels emotionally and narratively safer. If you listen to advocates, you can point to specific communities and say, “We’re answering your needs.” But it also turns discovery into a waiting game—something depends on who can advocate loudly enough and who can fund a bespoke program.
The AI-first route is less intuitive to donors and harder to market, and that’s where the strategy begins to look like a philosophy. What this really suggests is that the organization is trying to defeat randomness, not just speed up research.
One thing that immediately stands out is how radical “disease agnostic” becomes once money enters the room. Philanthropy in rare disease often reflects personal grief and urgency; donors usually want direct, disease-specific outcomes. So a strategy that “might help many” can feel too indirect—even when the indirectness is the entire point.
Personally, I find that trade-off both painful and revealing. It shows how even well-meaning funding can unintentionally reinforce the same siloed approach that keeps rare diseases stuck in the margins.
End-to-end work, not publication theater
Every Cure doesn’t just aim to publish drug–disease matches. It emphasizes going further: validating, shepherding candidates toward trials, engaging with regulators, and ultimately getting to patients.
From my perspective, this is where so many “breakthrough” models quietly fail. The modern biomedical ecosystem is full of dashboards, papers, and promising leads that never fully cross the long gap between insight and treatment. The gap exists because validation costs money, requires expertise, and demands patience.
What makes their end-to-end ambition particularly interesting is that it treats discovery as only the first act. It implicitly acknowledges that the hard part isn’t only finding a hypothesis—it’s proving it in the messy, real world.
A detail I find especially interesting is the focus on feasibility: candidates must look medically promising and financially realistic to advance. Personally, I think this is the practical discipline rare-disease innovation often needs but rarely admits.
People usually misunderstand this as “resource limitation,” when it’s actually a design constraint that prevents pipeline paralysis. If you can’t realistically finish the job, “finding matches” becomes an academic exercise rather than a patient-centered mission.
AI speeds the search—but validation still decides the fate
The technology side is striking: they score roughly 4,000 existing drugs against thousands of diseases and potential matches, turning a task that previously took days into something closer to hours.
What makes this more than a headline is what it implies about the workflow. In my opinion, AI here functions less like a magical oracle and more like an aggressive triage tool. It reduces the search space so teams can spend human effort where it matters.
But here’s the deeper question: if AI can generate lists that are fast, does the system also have the patience and infrastructure to test wisely? The fact that they use a medical team to narrow leads matters. Speed without governance just produces noise.
From my perspective, the real advantage is not that AI “discovers cures overnight.” It’s that it changes the calendar. In rare diseases, calendar time is biological time.
And what many people don’t realize is how small scheduling differences can reshape outcomes. Waiting months longer for a candidate can mean losing momentum in recruiting studies, recruiting patients, and completing preclinical steps.
Repurposing isn’t a replacement—it’s a parallel track
A crucial nuance in their stance: repurposing shouldn’t replace novel drug development. Some conditions truly require brand-new mechanisms, brand-new compounds, or advanced targeting.
Personally, I think the most responsible innovation strategy is portfolio thinking—building multiple paths and letting different diseases benefit from different kinds of invention. Treating repurposing as the only answer is as naïve as treating novel drugs as the only answer.
The argument for parallel tracks also has a financial rationale. Developing a brand-new drug can cost billions and take a decade or more, while repurposing can avoid parts of that long journey.
From my perspective, this debate often becomes ideological, but it should be operational. The right question isn’t “repurposing versus novel.” It’s “which diseases can be helped sooner, and what does the system do to make that happen?”
The uncomfortable problem: old drugs can be unprofitable to revive
Even when repurposing looks scientifically plausible, there’s an economic trap. If a drug is off-patent or barely worthwhile, manufacturers may not want to produce it. Sometimes the act of restarting supply can cost more than they can earn.
What this really suggests is that medicine isn’t governed solely by biology; it’s governed by supply chains and incentives. Personally, I think we understate how much the “last mile” depends on business decisions we never see.
If you take a step back and think about it, this is the hidden reason repurposing struggles: discovery is only half the battle. The other half is whether any actor in the market is motivated to manufacture, scale, and sponsor the next regulatory steps.
This raises a deeper question for me: are we designing health systems that assume profits will naturally align with public need? In rare disease, that assumption routinely collapses.
Regulatory reality: the sponsor model doesn’t fit the mission
Every Cure also runs into the FDA approval environment, which is still built around a traditional sponsor model—usually the people who made or market the drug and therefore have direct commercial interest.
Personally, I think this mismatch is one of the most telling parts of the story. Regulators are not villains; they’re operating within frameworks meant for a different ecosystem. But if the framework expects a sponsor with incentives, then non-profit and repurposing models face an uphill credibility battle.
What many people don’t realize is that this affects more than paperwork. It affects insurance pathways, clinician confidence, and overall awareness of whether a treatment is “real” in the eyes of the broader medical system.
So their emphasis on educating doctors makes sense. Even if FDA approval isn’t always required for off-label prescribing, approvals function like social proof as much as legal permission.
Funding choices reveal what the strategy truly values
Their early fundraising experience is a brutal but important detail: they couldn’t raise money in year one, and many donors wanted disease-specific targeting rather than disease-agnostic exploration.
From my perspective, that’s not just a fundraising anecdote—it’s a window into how the system rewards narratives. Donors often want a storyline with a single protagonist: “our disease.” The disease-agnostic model is harder to imagine, so it gets dismissed.
Personally, I think the fact they turned down money in the early years demonstrates an insistence on scientific integrity over emotional convenience. It’s the kind of decision that looks “risky” to outsiders but may be exactly what prevents mission drift.
And when credible funders do back this approach—philanthropic groups, major research investors, and federal funding streams—it suggests the ecosystem is slowly recognizing that discovery needs infrastructure, not just goodwill.
The Bachmann-Bupp example: proof of concept, not a finish line
Their work on Bachmann-Bupp syndrome illustrates the model’s potential. By identifying an older drug that may inhibit a protein involved in the disease, they report that a handful of patients—especially children—show meaningful improvement.
Personally, I’m encouraged by this kind of outcome because it aligns with why repurposing is so compelling: you can sometimes translate existing mechanisms into new clinical promises faster than starting from scratch.
But I also think it’s essential not to overread small numbers. In early-stage repurposing, the signal is hopeful; the next question is whether it holds up across broader cohorts and rigorous controls.
What this really suggests is that Every Cure’s pipeline is designed to take early promise seriously and move it forward. The editorial lesson here is that modern rare disease progress often comes from disciplined escalation, not from one miraculous result.
Where this could go next
If AI-based triage becomes a stable engine, the future could look like an always-on “searchlight” over existing medicine. Instead of waiting for a specific disease advocate to trigger discovery, the system continuously identifies candidates worth testing.
Personally, I think the bigger shift is cultural: it reframes repurposing as an ongoing commitment rather than a series of occasional projects. And once it becomes continuous, it starts to create institutional knowledge—about which targets transfer, which outcomes generalize, and which pipelines reliably reach patients.
The risk, of course, is that speed and optimism could outrun evidence. That’s why their end-to-end emphasis matters: without the validation and regulatory education, a fast pipeline can become a fast disappointment.
The takeaway nobody can ignore
Rare disease families are used to hearing that their conditions are too small, too complex, or too financially unattractive. Every Cure is effectively arguing against that fatalism by building a system that looks anyway.
From my perspective, that’s the most provocative idea here: the question isn’t whether cures exist—it’s whether we’ve arranged incentives and infrastructure to search for them.
And if you take a step back and think about it, the real revolution may not be AI. It may be the refusal to accept “not profitable” as the final word on “not treatable.”
