This is a sharp concern, and the research gives us a clear picture. Here's the honest breakdown:
Can Users Access Paywalled Content Through OpenEvidence?
The short answer: Yes - but indirectly, and by design.
When a publisher/society licenses their content to OpenEvidence, physicians can get AI-synthesized answers drawn from that paywalled content - without holding a personal subscription to the journal. OpenEvidence itself is free to any US physician with an NPI number (National Provider Identifier). So technically, a doctor who never paid for a NEJM or Cochrane subscription can now get answers grounded in that content.
The "Disintermediation" Risk You're Pointing At
This is a real tension in the industry. The concern is:
- A society journal charges members $500/year partly because members need access to the content
- Physicians can now get the clinical answers from that content via OpenEvidence for free
- Result: membership/subscription value proposition weakens, and society revenue could erode
This mirrors the broader "AI disintermediation" pattern seen across media and publishing.
How Publishers Are Framing It (and Why They're Still Signing)
The Sequoia Capital interview with OpenEvidence founder Daniel Nadler reveals a key insight: OpenEvidence sends millions of physician visits back to journal pages as cited sources. The model is:
AI answer + citation link → physician clicks through to the original article → journal gets traffic and engagement
So publishers are betting that OpenEvidence acts as a discovery and referral engine, not a substitute. The logic: a physician who would never have found or read a niche Cochrane review now encounters it, clicks the citation, and the journal gets a verified, engaged reader.
But the Risk Is Real - Especially for Societies
For medical societies specifically, the calculus is more complicated than for commercial publishers:
| Factor | Commercial Publisher | Medical Society |
|---|
| Revenue model | Mostly licensing/subscriptions | Membership fees + journal access bundled |
| Content exclusivity | Strong (large back-catalogue) | Moderate (guidelines, standards of care) |
| AI cannibalization risk | Lower (already diversifying revenue) | Higher (membership access is the core value prop) |
| Upside from AI partnership | Licensing fees + reach | Mission impact + some licensing revenue |
Societies like the American Diabetes Association, American College of Cardiology, and AAO-HNSF have already partnered with OpenEvidence - so they've evidently decided the reach and mission benefit outweighs the membership access risk. But this is a legitimate concern they'd need to address in how they structure the deal (e.g. linking full-text access to members only, while OpenEvidence only surfaces synthesized answers with citations).
The Key Structural Question for Prospecting
When approaching a society, the honest framing should address this directly:
- OpenEvidence surfaces synthesized answers with citations, not full article text
- Full-text access still requires clicking through to the journal (which can remain paywalled/member-only)
- The partnership can be structured so that full text is only unlocked for members, while non-members see the AI answer + a prompt to join/subscribe for full access
- This could actually increase membership conversions by exposing non-members to the value of the society's content at the moment they need it most
This "freemium gateway" framing - where OpenEvidence is the top-of-funnel and membership is the unlock for full text - would be a strong angle in publisher/society prospecting conversations.