Editorial Policy

How this site is operated, how content is produced, and what quality controls we apply — especially for health-sensitive supplement topics.

Site operations

dietarysupplement.ai is an AI-assisted publication. The infrastructure, editorial system, and quality controls are designed and operated with the help of Claude AI (Anthropic). News articles are drafted with the help of Google Gemini. Every article passes through human editorial review before it goes live.

How news articles are produced

Our daily news section follows a strict pipeline:

  1. Topic discovery. Daily trending searches from Google Trends are first checked for a credible supplement angle (a real ingredient, recall, FDA action, or clinical study). If nothing fits, a Gemini search-grounded fallback finds a recent supplement-specific story instead. Stories about prescription drugs, diet trends, or general wellness are explicitly rejected.
  2. Drafting. Google Gemini drafts the article body, summary, headline, and structured metadata using sources surfaced during topic discovery. The model is instructed to attribute sources inline and end with a "What this means for consumers" practical-takeaways section.
  3. Cover image. Gemini's image model generates a photorealistic editorial cover image for each article (1200×675).
  4. Bilingual translation. Each article is translated into Simplified Chinese before publication, so EN and ZH versions go live together.
  5. Editorial review. Drafts wait in our admin queue. An editor opens each one, reviews title, body, sources, and cover, and clicks Publish. Drafts are never published automatically — the cron-driven pipeline only produces drafts.
  6. Publication and indexing. Approved articles are published to /news/ (English) and /zh/news/ (Chinese) simultaneously. The Google News sitemap, archive sitemap, and Atom feeds are regenerated; IndexNow notifications are dispatched to participating search engines.

How older content (Articles / Ingredients / Products) is produced

Articles, ingredient summaries, and product pages predate the current system and follow a more traditional research-and-edit process. They are progressively being upgraded to align with the standards below.

Research standards

For evidence-grounded content we begin with authoritative sources such as NIH Office of Dietary Supplements fact sheets, FDA safety communications, PubMed-indexed clinical trials, systematic reviews, meta-analyses, professional society guidance, and product-independent testing resources when available.

Evidence grading

GradeMeaningTypical evidence
StrongConsistent human evidence or established essential-nutrient roleMultiple trials, guidelines, DRIs, or accepted clinical use
ModeratePromising but not definitiveSmall RCTs, mixed meta-analyses, or condition-specific evidence
PreliminaryEarly, indirect, or mechanisticAnimal, in vitro, biomarker, or limited pilot data

Use of AI tools

This site openly uses large language models. Specifically:

AI output is treated as a draft, not a source. Medical claims must be checked against primary or authoritative references. The human editor is responsible for what is actually published.

Corrections and updates

We correct factual errors, outdated safety information, broken references, and claim-strength problems as soon as they are identified. To report an error, email editorial@dietarysupplement.ai.

Funding and conflicts of interest

Article topics are selected based on news relevance and search demand, not paid placement. Affiliate links, when used elsewhere on the site, are disclosed on the relevant page.

Medical disclaimer: dietarysupplement.ai publishes educational information only. Our content is not medical advice and is not a substitute for diagnosis, treatment, or care from a qualified healthcare professional. These statements have not been evaluated by the Food and Drug Administration. Supplements are not intended to diagnose, treat, cure, or prevent any disease.