Artificial Intelligence Use Policy.
Where artificial intelligence is used in the laboratory's processes and the limits of that use.
Updated — First editorial cycle — 2026
General principle
Artificial intelligence tools may support technical steps of investigative work, but they do not replace reporting, editorial judgement or human responsibility. Every PLOT. publication is conceived, verified and signed by people.
Permitted uses
- Transcription of interviews and audio, always checked against the original material.
- Optical character recognition on scanned documents.
- Organising, classifying and searching large document collections.
- Support in exploring databases and writing analysis code, with review of the results.
- Grammar review and clarity suggestions on text written by people.
Forbidden uses
- Generating journalistic text published as a story.
- Creating photorealistic images that could be mistaken for documentary record.
- Producing quotes, sources, documents or data.
- Editorial decisions about publishing, framing or omitting information.
- Sending confidential material, sensitive personal data or information capable of identifying protected sources to third-party services.
Verification
Results produced with the support of artificial intelligence are treated as hypotheses, never as evidence. Every piece of information obtained this way is checked against the primary source before it is included in a publication.
Disclosure to the reader
When an artificial intelligence tool plays a relevant role in producing a published element — a large-scale classification, an analytical model, an admittedly synthetic illustration — this is explicitly disclosed in the publication's methodology.
Responsibility
Editorial responsibility for any published content lies entirely with PLOT. and whoever signs the investigation. The use of technical tools does not reduce that responsibility.