AI-Powered Mapping of Psychedelic Clinical Trials: Global Registry Insights

AI Enables Systematic Mapping of Psychedelic Clinical Trials

A 2026 cross-sectional analysis demonstrates that artificial intelligence (AI) can systematically and reproducibly map the global landscape of psychedelic clinical trials by leveraging large language models (LLMs) connected to the ClinicalTrials.gov registry application programming interface (API). This approach, described by the study's authors (source), replaces labor-intensive manual synthesis with protocol-driven, field-specific queries, offering a transparent and scalable method for tracking the rapidly evolving field.

The researchers used Anthropic's Claude LLM to retrieve and structure data on interventional trials for classic psychedelics (psilocybin, MDMA, LSD, DMT/5-MeO-DMT, ayahuasca, ibogaine) as well as ketamine/esketamine comparators. The method involved two querying strategies: open-ended natural language and a more rigorous, documented protocol of explicit registry field expressions. The latter approach produced more accurate and reproducible counts, revealing substantial underestimation by conversational querying alone—a key insight not widely recognized in prior registry analyses.

Key Findings: Compound Focus and Indication Trends

AI-driven registry queries reveal that psilocybin and MDMA dominate the current clinical trial landscape, with a pronounced focus on depression and post-traumatic stress disorder (PTSD). Of 412 core-set psychedelic trials (excluding withdrawn records), psilocybin accounted for 244 (59%), with 181 enrolling patients; MDMA followed with 99 trials (47 in patients). LSD, DMT/5-MeO-DMT, ayahuasca, and ibogaine were far less represented, with 39, 36, 6, and 4 trials, respectively.

• Psilocybin trials primarily target depression (74 of 181 patient trials), with 63 phase 2/3 studies in this indication. Over half (56%) used a depression rating scale—most often the Montgomery-Åsberg Depression Rating Scale (MADRS)—as the primary endpoint.
• MDMA trials are concentrated on PTSD (35 of 47 patient trials), reflecting the compound's advanced development for this indication.
• Ketamine and esketamine, reported separately, account for 300 trials, with depression as the leading indication (237 trials).

The study also highlights trial design features: 41% of psilocybin depression trials excluded adults over 65, and none enrolled adolescents, indicating a persistent age gap in research. Notably, only 24 core-set trials were registered as phase 3—13 for psilocybin, with planned enrollment exceeding 1,900 participants, compared to much smaller numbers for MDMA and LSD.

Geographic and Methodological Disparities

AI-powered registry mapping exposes a stark geographic imbalance in psychedelic clinical research: only five patient trials (2%) had a site in a low- or middle-income country (LMIC), underscoring persistent global inequities in research access and capacity. This finding is particularly salient for policymakers and funders, as it quantifies the often-cited but rarely enumerated exclusion of LMICs from the psychedelic research agenda.

Methodologically, the study demonstrates that protocol-driven, field-specific AI queries yield more comprehensive and reproducible trial counts than open-ended or conversational LLM approaches. For some compounds, conversational querying underestimated the number of trials by several-fold, a failure mode not previously highlighted in registry-based reviews. The authors emphasize that while AI-assisted mapping complements traditional systematic reviews, it does not replace manual validation, especially for nuanced trial characteristics not captured in structured registry fields.

Policy, Research, and Industry Implications

The ability to rapidly synthesize the global trial landscape has several implications for researchers, sponsors, and regulators. For researchers, AI-driven registry mapping can identify gaps in compound, indication, or population coverage, guiding new study proposals or meta-analyses. For sponsors and industry stakeholders, the data inform strategic decisions about compound selection, trial design, and market differentiation. Regulators and policymakers gain a transparent overview of the field's maturity, geographic reach, and methodological rigor, supporting evidence-based policy development and resource allocation.

Importantly, the study’s protocol-driven approach sets a new standard for registry-based surveillance, improving transparency and reproducibility over manual or ad hoc reviews. However, the authors caution that registry data may not fully capture unpublished, withdrawn, or non-registered studies, and that structured fields may omit critical information about trial conduct or outcomes.

Risks, Unknowns, and the Path Forward

AI-assisted registry mapping introduces new risks and unknowns, including potential errors from incomplete or inconsistently coded registry data, and limitations in LLM comprehension of complex trial protocols. The accuracy of AI-driven synthesis remains to be validated against traditional systematic review, and there is a risk of overreliance on registry data that may lag behind real-world trial activity.

Another non-obvious implication is that as AI-driven mapping becomes standard, the quality and completeness of registry entries will become a critical determinant of field-wide visibility and funding. This may incentivize sponsors to improve registry reporting, but could also widen disparities if resource-limited investigators are unable to meet evolving data standards.

Looking ahead, the integration of AI with global trial registries promises to accelerate evidence synthesis and strategic planning in psychedelic science. However, robust validation and attention to equity will be essential to realize these benefits without exacerbating existing gaps.

How we research: This article was written and reviewed by Dr. Alex R. Bennett, PhD (neuroscience, clinical trial methodology), on 2026-10-16. Primary data and methodology are drawn directly from the published study and ClinicalTrials.gov registry records.