Trust and accountability
These standards explain how 221B Face Search researches, reviews, labels, updates, and corrects product guidance about public-source face-search workflows.
221B Face Search publishes product guidance and educational material about public-web face search, source review, catfish checks, responsible OSINT, photo misuse, and online safety. We write for readers who need a practical review process, not an automated identity verdict.
We prefer primary sources such as official product documentation, provider documentation, public policies, standards, and original research. When a secondary source is useful, we check its evidence and make clear what is reported, what is observed, and what is our interpretation.
We describe work as tested, reviewed, or observed first-hand only when we performed that work. When first-hand testing matters, we aim to state the method, relevant date, sample constraints, and limitations. Reading documentation alone is not described as hands-on testing.
A candidate face-search result is a lead for manual review, not proof of identity, ownership, intent, fraud, or safety. A match-quality score is a ranking signal, not a probability. No result is inconclusive and does not prove that an image or identity claim is genuine.
221B Face Search is our own product. When we compare it with another service, we disclose that relationship, define criteria before drawing conclusions, and avoid presenting commercial preference as independent fact. We do not hide paid placement or compensation.
AI tools may assist with structure, drafting, or consistency checks. Di Tang remains responsible for source verification, material claims, safety boundaries, and the published result. AI assistance does not replace review or create first-hand experience that did not occur.
We review material claims when product behaviour, pricing, provider documentation, public policy, or cited evidence changes. Material corrections should update the page date and, where useful, include a visible correction note. Readers can report an issue through our corrections process.
We encourage lawful, proportionate use of public information and collection of only what is needed for review. We do not position similarity results as a basis for high-impact decisions, harassment, surveillance, or conclusions about protected or highly sensitive traits.
Send the exact page URL, the statement you believe is inaccurate, and a reliable supporting source. Do not send passwords, identity documents, payment-card details, or private images.
