AI-Use Policy

We use AI tools in our workflow, and we think readers deserve to know exactly how. Used carelessly, AI is a machine for producing confident, unsourced, subtly-wrong medical text — the precise opposite of what this site is for. We use it as an assistant under human control, never as an unaccountable author.

Where we use AI

• Drafting and restructuring prose from human-gathered sources.
• Surfacing candidate literature for a human to evaluate.
• Consistency checks — flagging missing fields, dead links, or duplicated coverage.
• Formatting structured data into our content model.

Where a human is always required

• Evidence grades. Every A–U grade is a human editorial judgment, not a model output taken on faith.
• Load-bearing facts. Numbers, regulatory status, and safety statements are verified by a human against primary sources.
• Citations. A source is only cited after a human confirms it exists and says what we claim it says.
• Publication. Nothing evidence-sensitive publishes without human review.

Lines we do not cross

• We do not publish AI-generated citations without verifying them — invented references are a known failure mode and we treat them as disqualifying.
• We do not let a model assign or move an evidence grade on its own.
• We do not present AI-written text as the work of a named human expert.
• We do not use AI to fabricate reviewers, credentials, or authority.

Accountability

Regardless of the tools involved, Synptide is accountable for every published claim. If a page is wrong, that is on us, and ourcorrections policy applies exactly as it would for any other error.