AI Disclosure
Last updated: May 2026
The Fetch incorporates AI tools into its editorial workflow. This page provides a clear summary of where AI fits into our process, how it supports our work, and where human editing remains essential.
Our position on AI
We use AI as a tool for research and writing efficiency, but never for key editorial decisions. All articles are assigned, sourced, and reviewed by humans, ensuring AI assists but does not direct or decide content.
How AI is used in our pipeline
Our editorial pipeline has four stages. AI plays a role in three of them, always within bounded, source-grounded constraints.
Stage 1: Source curation. An AI assistant helps identify authoritative sources on a given topic — peer-reviewed papers, veterinary teaching references, regulatory documents, and similar. The AI suggests sources; the editor verifies each suggested source is real, accessible, and authoritative before it enters the pipeline.
Stage 2: Source-grounded extraction. The verified sources are ingested into a retrieval-augmented generation environment. Editorial questions are used to elicit factual answers, with citations to the source material.
Stage 3: Drafting. A human writer, together with an AI writing assistant, co-drafts an article using only the source-grounded findings from Stage 2.
Stage 4: Editorial review. A human editor reviews every draft before publication. The review checks that claims are accurately represented from their underlying sources, that the article serves the reader’s interests, that the tone is appropriate, and that nothing important has been lost in the drafting process. The editor revises, rewrites, or rejects drafts at this stage.
What AI is not used for
AI is not used to make editorial judgments about what topics to cover, what conclusions an article should reach, or which sources to trust. Those decisions are made by humans.
AI is not used to fabricate quotes, attribute statements to people, or invent information not present in the source material. The pipeline is structured to prevent this.
AI-generated images are clearly identified when used. AI-generated text without the source-grounded pipeline described above is never published.
Why this matters
AI tools can be powerful, but they can also fabricate convincing falsehoods if used without constraints. Pet health is a domain where inaccurate information has real consequences for animal welfare. We have built our editorial pipeline specifically to harness the speed and breadth that AI offers while structurally constraining its known failure modes — through source-grounding, citation discipline, and human editorial review at the end of every article.
Contact
If you have questions about how AI is used in a specific article or in our editorial process generally, contact us.