AI Output & Research Disclaimer
Effective date: October 1, 2026
SynthGap uses artificial intelligence to speed up parts of a systematic review. This page explains what the AI does, what it can get wrong, and what you remain responsible for. It forms part of our Terms of Service.
In short: SynthGap drafts; you decide and you are accountable. Treat every AI-generated screening decision, extracted number, risk-of-bias answer, topic suggestion and paragraph of text as a draft to verify against the source, never as a finished result.
1. What the AI does at each stage
| Stage | What AI does | What code computes (no AI) | What you must verify |
|---|---|---|---|
| Topic discovery | Proposes candidate meta-analysis topics from retrieved PubMed records, checks them against an index of existing systematic reviews, and writes a summary with a PICO framework | Retrieval and similarity search over our PubMed index and review index | That the topic is genuinely novel and feasible. Search other databases and PROSPERO before you commit |
| Protocol & PICO | Drafts a protocol, PICO question and eligibility criteria for the topic you pick | None | All of it. Lock the protocol only when you are satisfied |
| Search strategy | Drafts Boolean search strategies | None | Run and refine them in the real databases (for example PubMed, Embase, CENTRAL). Our index is not a complete database |
| Title/abstract and full-text screening | Recommends include or exclude, with reasons | None | Every decision. Your confirmation, not the AI's suggestion, is what counts |
| Data extraction | Extracts study data in two independent AI passes and highlights disagreements | The comparison of the two passes | Every extracted value against the source, especially numbers, units and denominators |
| Risk of bias | Picks an appropriate tool (RoB 2, ROBINS-I or Newcastle-Ottawa) and answers its signalling questions | The overall judgment and rating are recalculated in code from the answers | The answers and the tool choice. These are judgments, not facts |
| Evidence synthesis | None | Random-effects pooling, heterogeneity, subgroup, sensitivity and publication-bias analyses, and certainty-of-evidence ratings | The input data, the model choices, and the certainty-domain judgments (which you can override) |
| PRISMA 2020 flow and checklist | None | Flow counts and checklist item status, computed from the records in your review | That your records are accurate and that items marked "reported" really are |
| Manuscript write-up | Drafts the manuscript sections | None | Everything. Rewrite in your own words, and check every claim and citation against your data |
2. What can go wrong
- The AI can be wrong. Language models can misread a paper, invent details, misclassify a study, or state something confidently that isn't in the source. Extracted numbers are the highest-risk output.
- Coverage is incomplete. Our indexes are built from PubMed and PubMed Central. They don't include every database, trial registry or the grey literature, so the Service can miss relevant studies and existing reviews. "Novel" means "not found in our index," not "nobody has done this."
- Full text is partial. Open-access full text is available for only part of the literature. For the rest you can upload files. Text extraction from PDFs can drop tables and figures, and scanned images can't be read at all.
- Judgments are subjective. Risk-of-bias and certainty-of-evidence ratings involve judgment. The Service proposes them; qualified reviewers should decide them.
- Statistics depend on inputs. Pooled estimates are only as good as the data entered and the model chosen. With few studies, heterogeneity and publication-bias tests are unreliable.
- Results can vary. Running the same task twice may give different AI outputs.
- Free plan. Each Free-plan run returns a single topic, chosen from the lower-confidence end of what the run found. The strongest candidates are shown only on paid plans.
3. Your responsibilities
- Verify every AI output against the source before relying on it.
- Follow your field's methodological standards. Most (for example the Cochrane Handbook and JBI guidance) expect independent human reviewers for screening and data extraction. The Service is designed to support that process, not replace it.
- Check that every citation, PMID and number is real and says what the draft claims.
- Register your protocol where appropriate (for example PROSPERO or OSF) and follow the reporting guideline for your review type.
- Follow the AI policies of your journal, funder and institution, and disclose your use of AI tools as they require. Under ICMJE recommendations, for example, AI cannot be listed as an author and its use should be disclosed.
- Keep your own audit trail of decisions and changes.
4. Not medical advice
Nothing the Service produces is medical advice or a clinical recommendation for any patient, clinician or other person. Outputs are research-synthesis drafts intended for qualified researchers exercising independent professional judgment. Don't make or change treatment decisions based on them.
5. Not authorship; no guarantee of acceptance
SynthGap is a tool, not an author, and takes no responsibility for authorship criteria, journal or funder requirements, research-integrity standards, or whether a manuscript will be accepted. You alone are responsible for the accuracy, integrity and compliance of anything you publish or submit.
6. How your data is handled
To generate outputs, parts of your content are sent to the AI and infrastructure providers on our Subprocessors page. We don't use your content to train AI models. Don't upload information that identifies a patient or other individual. See our Privacy Policy.
7. Found a problem?
Please tell us at support@synthgap.com: what you ran, what you expected, and what the Service produced. We can't fix what we don't know about.