In late 2025, CEPI and PATH began working on a simple question with big implications for outbreak response: Could market-ready artificial intelligence (AI) help teams draft high-quality, compliant clinical trial protocols faster, without sacrificing scientific rigor or usability?
For CEPI’s 100 Days Mission, shortening protocol timelines could mean starting trials sooner. For PATH, it was an opportunity to test emerging tools against real-world clinical research requirements.
The partnership agreed early on not to begin with a pilot. Instead, CEPI and PATH structured this project as a decision-support effort: define what “good” looks like, invite the market to respond, and then test claims in a way that would withstand scrutiny from clinical, technical, and regulatory perspectives.
Starting with evidence
The work began with groundwork that often gets skipped in fast-moving technology projects. PATH analyzed peer-reviewed and gray literature (2019–2025) to understand how AI-enabled documentation tools have been evaluated and where the evidence is thin for protocol-specific use.
In parallel, CEPI and PATH mapped the real workflow for protocol and essential document development, aligning it with the International Council for Harmonisation (ICH) E6 (R3) Good Clinical Practice guideline, a globally recognized standard that helps ensure clinical trials are scientifically sound, ethically conducted, and acceptable to regulators across multiple jurisdictions. We also aligned with CEPI expectations and common regulatory requirements, including U.S Food and Drug Administration (FDA) Investigational New Drug (IND) requirements and National Institutes of Health (NIH)-funded trial needs.
By grounding the workflow in these internationally accepted standards and regulatory expectations, the team ensured that the evaluation framework reflected the practical realities of developing protocols that can support high-quality, compliant clinical research. That combination of evidence review and process mapping became the backbone of a practical scorecard for judging tools on quality, compliance, usability, and implementation readiness.
From the outset, CEPI and PATH assessed tools across five lenses: technical performance, regulatory compliance, document quality, usability in real trial workflows, and implementation readiness (including security and suitability for low-resource settings).
With the evaluation approach in place, the project moved from theory to the market.
“Our goal was not automation for its own sake, but responsible acceleration. AI can support early drafting, but expert oversight and governance must remain central.”— Linda Taylor, Technical Program Manager, PATH
Opening the door to the market
In December 2025, CEPI and PATH published a global Expression of Interest, inviting market-ready “AI-for-protocols” solutions and requesting comparable information across vendors about tool capabilities, evidence of use, governance, and practical constraints such as cost and licensing.
To support broad participation, the Expression of Interest was shared through CEPI and PATH channels and accompanied by an informational webinar. When vendors requested more time, the submission window was extended to ensure a competitive and diverse response.
By the final deadline, ten proposals had been received, enough to confirm that the AI-for-protocols market is not only active but also varied in maturity and evidence. Participating vendors included Verily, Emmes Group-Veridix, Risklick, Kitsa, EpiDriver, Synterex, Izertis, and Saama, among others.
Note: This article describes participation in the process; it does not rank or endorse specific products.
Submissions covered a range of approaches, such as AI-assisted drafting, structured templates, and consistency checks across documents, along with information on prior use, governance, and implementation readiness.
“What we saw was a rapidly evolving landscape. There is real innovation, but also wide variation in maturity, evidence, and readiness for outbreak use.”— Sue Wairimu, Associate Product Manager, Living Labs Initative, PATH
Selecting and testing tools
With proposals in hand, CEPI and PATH convened a multidisciplinary technical evaluation committee in February 2026. The committee brought together experience in software maturity assessment, AI-enabled documentation tools, clinical trial operations, and regulatory and implementation considerations.
Members scored submissions independently against the shared scorecard, then met to reconcile results and document the basis for short listing.
This step clarified which solutions were strong candidates for further consideration, which showed promise but needed maturity, and which did not meet CEPI’s requirements for outbreak-relevant protocol development.
Written proposals can only go so far, so CEPI and PATH added a scenario-based test. Short-listed vendors were given the same outbreak-relevant scenario and asked to produce a draft protocol, an informed consent form, and one additional document within defined constraints.
Reviewers assessed outputs for completeness, quality, alignment with ICH expectations, and practical usability in real workflows. The exercise revealed where tools genuinely accelerated early drafting and where gaps still required substantial expert intervention, providing evidence for next steps.
Learning and next steps
The process surfaced a clear picture of an emerging market: Multiple tools can support faster early drafting, but maturity and evidence vary widely. It also reinforced that hands-on testing is essential: Scenario-based document generation provided practical insights.
Finally, the project produced a reusable evaluation method that CEPI can apply as AI tools evolve and as expectations for responsible use become more defined.
Now, CEPI and PATH are using the results to plan a controlled pilot focused on measurable time savings, document quality, and strong governance. This means keeping human experts in the loop and aligning with regulators and ethics requirements.
This project, dubbed "AI4Protocols," underscored a practical takeaway for outbreak preparedness:
“AI is not a substitute for scientific and regulatory expertise but when used responsibly, it can help teams move faster and more consistently, especially in high-pressure outbreak settings.”— Roland Tschismarov, Clinical Development Science Lead at CEPI
This project highlights a broader shift in outbreak preparedness: combining digital innovation with strong governance to improve speed without compromising quality.
By bringing together CEPI’s mission-driven focus with PATH’s applied evaluation approach, the collaboration offers a replicable way to assess AI tools in high-stakes clinical research workflows.
As outbreaks become more frequent and complex, tools that responsibly accelerate clinical research could play a critical role in global response efforts.
CEPI and PATH will continue to share learnings as piloting plans progress and as the evidence base for AI in protocol development grows.
For more information: Learn about CEPI’s 100 Days Mission (https://cepi.net/100-days-mission)
