How we report

1. What we cover

We cover AI pet technology — the collars, the apps, the science, the startups, the business model. The criteria for what makes the coverage queue:

• Newsworthy. Something verifiable just happened — a product shipped, a study published, a company raised, a regulatory move landed.
• Underreported. The angle isn't already saturated by the press-release rewrite economy.
• Decision-relevant. The story helps a pet owner, a vet, an investor, or a fellow journalist make a better decision.

2. Sourcing

Every factual claim in a flagship piece traces to a primary source — a company press release, a regulatory filing, a peer-reviewed paper, a named on-record interview, or directly observed product behaviour. We cite inline and (on long pieces) list sources at the foot of the article.

When a number we'd like to report can't be verified, we say so. "Unverified by us" and "we've reached out and will update" are honest disclosures.

3. Fact-checking

Every flagship piece is read end-to-end by a second editor before publication. Specific numbers — prices, accuracy claims, ship dates, funding amounts, percentages — are checked against the primary source at the moment of publication and re-checked on any material update.

4. Attribution

Named-author pieces are written and signed by the listed author. "Editorial team" bylines are multi-author or include contributions where attribution to one person would be misleading.

We never use AI to draft editorial content. We do use AI for research support — summarizing long PDFs, scanning patent filings, translating Chinese-language sources — and the underlying source is checked before anything from it appears in a published piece.

5. Conflicts of interest

We hold no equity, paid consulting relationships, or family ties at any company we cover. If that ever changes we disclose it inline on every relevant story and update our disclosure page.

6. Free product units

We sometimes receive review units at no cost. When that happens we disclose it inline at the top of the relevant story. The unit goes back to the manufacturer at the end of testing, or — if the company declines its return — is donated and the recipient is named publicly.

7. What we don't do

• No sponsored editorial. Brands can't pay us to run a story, and brands can't pay us to influence framing. Newsletter sponsorships are clearly labelled and never touch editorial.
• No pre-publication review. We don't show stories to subjects for approval. We send fact-check questions; we don't send drafts.
• No listicle traps. No "10 Best AI Dog Collars" pieces written purely to push Amazon clicks. When we list products, the list reflects what we'd actually recommend.
• No fake first-person. If we haven't tested something hands-on, the piece says so.

8. Updates and corrections

Stories get updated when underlying facts change. Material updates (new verdict, new pricing, new product version) are dated and summarised at the top of the post. Minor edits aren't called out individually but the modified date moves.

If we get something wrong, we correct it visibly. A corrections note appears at the foot of the affected post explaining what was wrong, what changed, and when. We don't silently rewrite history. Report a correction to editor@animalcollar.ai with a link and a primary source.

9. Privacy of subjects

We don't publish personally-identifying information about consumers without consent. Sources who request "on background" treatment get it. Public figures and corporate spokespeople are quotable in their public capacity.

This page is versioned

When anything material changes on this page we publish a note in The Weekly Brief and date the change here. Last updated: May 27, 2026.