Scientific Accuracy and Usage Rights

PaperBanana helps you turn research ideas into visual drafts, but it is not a scientific reviewer, rights lawyer, journal editor, or institutional compliance tool.

PaperBanana does not verify scientific truth. Treat every generated figure as a draft that needs human review before academic use.

Scientific accuracy checklist

Before using a generated visual in a paper, poster, slide deck, grant, or class material, check:

• Scientific claims, mechanisms, and cause-and-effect arrows.
• Cell types, molecules, pathways, organisms, methods, and experimental steps.
• Labels, abbreviations, units, scale bars, and legends.
• Whether the figure matches your source data, protocol, or manuscript.
• Whether any visual simplification could mislead readers.

If the figure is close but needs local fixes, use Workspace Editor. If the structure is wrong, regenerate with a more explicit prompt.

Usage rights checklist

Before uploading or reusing material, confirm that you have permission to use it:

• Use your own data, sketches, diagrams, or materials you are allowed to process.
• Do not upload patient data, private IDs, confidential manuscripts, unpublished third-party figures, or restricted institutional material.
• Check whether your journal, conference, funder, or institution requires AI disclosure.
• Keep records of source images, prompts, edits, and exports when your workflow requires traceability.
• Avoid presenting generated or edited visuals as measured data unless they are directly backed by your actual data.

For product boundaries, read What PaperBanana is not for.

Before submission or presentation

1. Review the figure with the manuscript, poster text, or slide context next to it.
2. Ask a domain expert or coauthor to check important mechanisms and labels.
3. Export the final file and inspect it at the final print or screen size.
4. Confirm journal, conference, course, or institutional disclosure requirements.
5. Save the final source file and exported file together with your project materials.

When to ask a human expert

Ask a human expert before relying on a generated visual when it explains:

• A disease mechanism, treatment pathway, or clinical workflow.
• A causal model, statistical workflow, or experimental design.
• A figure that could affect scientific interpretation or reviewer confidence.
• A visual based on unpublished, restricted, or third-party material.

If something looks wrong after generation, use Troubleshooting before spending more credits.