“ChatGPT for Doctors”? What Health IT Leaders Should Know About OpenEvidence

CHUG Insight

“ChatGPT for Doctors”? What Health IT Leaders Should Know About OpenEvidence

Introduction

In November 2025, Healthcare Brew profiled OpenEvidence—an increasingly prominent entrant in the clinical AI landscape, often referred to as “ChatGPT for doctors.” As AI becomes more deeply embedded in healthcare operations, the OpenEvidence model offers a useful case study for CHUG members evaluating evidence access, clinical search tools, and AI governance across their organizations.

The company’s core value proposition is simple and timely: give clinicians instant answers rooted only in verified medical research—not the open internet. With more than $400 million raised across recent rounds and a $6 billion valuation, OpenEvidence is emerging as a significant player in the growing category of evidence-bound clinical AI.

Why Tools Like OpenEvidence Are Gaining Traction

1. The volume of medical research is exploding.
Guidelines, FDA approvals, PubMed publications, and specialty society updates change weekly.

2. Traditional search workflows are slow.
Textbooks, paywalled databases, and multi-tab browser searches take time clinicians don’t have.

At HLTH 2025, OpenEvidence Chief Medical Officer Travis Zack described the original goal as straightforward “information retrieval.” As generative AI advanced, the tool evolved into a platform capable of scanning trusted literature, extracting clinically relevant evidence, and summarizing it within seconds.

A Clinicians-Only Tool Built to Preserve the Patient–Provider Relationship

A defining feature of OpenEvidence is its clinician-only access model. Users must verify their identity through NPI and credential checks. Zack emphasizes that the platform intentionally avoids contributing to trends that sideline clinicians:

“Diagnosis and treatment should be between a physician and a patient. We don’t want to disintermediate the provider.”

This approach positions OpenEvidence firmly as a tool for clinicians—not instead of clinicians.

The Evidence-Bound Model: Where OpenEvidence Stands Apart

Unlike general-purpose language models trained on large swaths of the internet, OpenEvidence limits itself to curated, trusted literature, including:

JAMA Network
New England Journal of Medicine
National Comprehensive Cancer Network
PubMed
FDA
CDC
Other licensed and peer-reviewed sources

The platform updates in real time as new research is published, and every answer links back to the original papers—aligning with the long-standing medical principle of “trust, but verify.”

Jack Lindeman, SVP of Commercial Strategy at OpenEvidence, explains:

“OpenEvidence develops state-of-the-art AI search algorithms that find the exact guidelines, papers, reviews, etc. that address the user’s question, then applies a layer of conversational fluency.”

This design aims to reduce hallucinations and improve transparency—two essential factors in responsible clinical AI adoption.

How Clinicians Are Using the Tool Today

Examples highlighted in the Healthcare Brew report show diverse applications:

Anesthesiology: Reviewing new cardiac protocols or unfamiliar medications.
Surgery: Comparing treatment pathways before complex cases.
Hospital medicine: Quickly confirming evidence for uncommon conditions.

Historically, clinicians relied on textbooks, journal sites, and multi-database searches. OpenEvidence consolidates this workflow into one fast interface—saving meaningful time in clinical decision-making.

Does AI Create Clinical Deskilling—or Clinical Acceleration?

Concerns about AI-enabled “clinical deskilling” have gained attention, but some experts see it differently. Mayo Clinic surgeon Antonio Forte argues:

“Now they will just have more information at their disposal at a faster speed than before.”

Rather than replacing medical reasoning, tools like OpenEvidence may augment it—delivering evidence clinicians previously accessed through slower, fragmented workflows.

Benefits and Trade-Offs of Evidence-Bound AI

For health IT leaders, the OpenEvidence approach surfaces several strategic considerations that go beyond “trustworthy sources.”

Benefits:

Stronger governance posture: A constrained, fully licensed corpus simplifies conversations with compliance, risk, and legal teams.
Clearer auditability: Being able to click through to every underlying article supports chart documentation, peer review, and quality improvement work.
Better alignment with clinical culture: Clinicians are accustomed to guidelines, pathways, and primary literature; OpenEvidence mirrors how they already think about evidence.

Trade-offs and gaps to watch:

Lag behind real-world practice patterns: Published literature often trails emerging bedside practice, local protocols, and operational realities.
Limited representation of diverse populations: If the underlying studies underrepresent certain demographic or clinical groups, the AI’s guidance may inherit that bias.
Fragmented documentation: If evidence queries live in a browser while clinical documentation lives in the EHR, organizations need clear guidance on how decisions are captured and shared.

Understanding these benefits and trade-offs helps organizations decide where an evidence-bound tool fits in their broader AI strategy, and where complementary tools (such as real-world data analytics or local pathway engines) may still be needed.

Questions Health IT Leaders Should Ask Vendors

Whether evaluating OpenEvidence or any similar AI-powered evidence tool, CHUG members can use a common set of questions to guide due diligence:

Corpus and licensing: What sources are included and excluded? How are licenses structured? How often is the corpus refreshed?
Explainability: Can clinicians always see which articles or guidelines informed a given answer?
Governance and controls: Are there role-based permissions, usage analytics, and administrative controls for clinical leaders?
Integration model: How can the tool be integrated with existing clinical systems, SSO, and identity management so it fits naturally into existing workflows?
Audit and risk: What logs are available if an answer is questioned later? How are safety issues reported and addressed?
Data handling: What PHI, if any, is sent to the vendor? How is it protected, retained, or de-identified?

These questions move the conversation from “Is the AI accurate?” to “Is this system safe, governable, and aligned with how our organization practices medicine?”

Implications for Health IT, EHR Teams, and CHUG Members

The OpenEvidence story also highlights broader trends that affect healthcare organizations of all sizes:

1. Demand for domain-specific AI tools.
Clinicians increasingly expect AI systems that are purpose-built for medicine, not repurposed from consumer use cases.

2. Rising expectations for traceability and auditability.
Boards, regulators, and quality committees are asking not just what AI recommends, but how it arrived there and what evidence it used.

3. Pressure on EHR and clinical systems to integrate AI search.
Regardless of which EMR a site uses, there is growing demand to surface evidence at the point of care instead of in separate, disconnected tools.

4. Expanded AI governance and compliance responsibilities.
Health IT teams must partner with clinical leaders to define safe use policies, evaluate vendors, validate data provenance, and train staff on appropriate use of AI in care delivery.

AI is not replacing clinicians; it is reshaping how information flows to them and how quickly evidence can inform everyday decisions.

Conclusion

OpenEvidence illustrates the broader evolution toward evidence-bound, clinician-first artificial intelligence. As the healthcare AI market accelerates toward an expected $187.7 billion by 2030, health IT leaders across all systems and EHR environments must evaluate how tools like this fit into clinical workflows, governance frameworks, and evidence access strategies.

Regardless of an organization’s size, infrastructure, or chosen EMR, one thing is clear: Clinicians will increasingly expect fast, trustworthy, and transparent access to medical evidence. IT teams and operational leaders are uniquely positioned to guide responsible adoption, ensure data provenance, and help shape the future of clinical AI inside their organizations.

Sources referenced: Healthcare Brew (Cassie McGrath, Nov. 14, 2025); The New York Times funding report; Research Insights projections; HLTH 2025 interviews with OpenEvidence leadership.