Editorial Policy
The editorial rules that keep dramatic AI-doomsday questions evidence-led, transparent about uncertainty and free from fabricated claims.
The dramatic-hook rule
We may use language such as “AI doomsday,” “takeover” or “apocalypse” when it accurately describes the question people are asking. The answer underneath must not turn an uncertain scenario into a fact merely because fear improves clicks.
Evidence labels
We distinguish: controlled evaluation, where researchers deliberately construct a test; real-world incident, where a deployed system or attacker caused an external event; forecast, where someone estimates a future outcome; and scenario, where a hypothetical chain is explored.
Real people, real credentials
We do not use invented author biographies, qualifications, testimonials, laboratory results or case studies. Any name, credential or expert review shown on the site must belong to a real, verifiable person.
No panic commerce
Commercial pages may discuss multi-use preparedness products, but we do not fabricate shortages, fake deadlines, guaranteed threats or “buy before AI takes over” claims. Advertising and affiliate relationships must be clearly disclosed.
Corrections and updates
Fast-moving incident and research pages carry review dates. When an underlying fact changes materially, the page should be updated rather than silently leaving an outdated claim in place.