Most readers don't have time to fact-check every AI story that crosses their feed, and they shouldn't need specialized technical knowledge to avoid being misled by one. A handful of quick, non-technical questions catch most of the weak reporting on this beat before it wastes your attention.
Is the source the company itself, or someone independent?
A claim sourced entirely to the company it's about - a press release, a company blog post, an executive's own social media - deserves less weight than one confirmed by an independent party. This doesn't make company-sourced claims false, just unverified.
Is there a specific number, or just a vibe?
"Massive," "unprecedented," and "game-changing" are opinions dressed as facts. A story with a specific, attributed figure - even an imperfect one - gives you something to evaluate. A story built entirely on adjectives doesn't.
Does the headline match the body?
AI headlines are especially prone to overstating a hedge buried three paragraphs down. If the headline claims a breakthrough and the body says "early testing suggests," trust the body.
Would this story exist without the company's cooperation?
A story that could only have been written with the subject's active participation - access, quotes, demos - is closer to sponsored content than reporting, whatever the byline says.
What's missing?
The most reliable tell is what a story doesn't mention: cost, failure rate, competitor context, or a timeline for when claims will be testable. Absence of these is itself information.
Why this is worth the ten seconds
None of these questions require domain expertise, and running through them takes less time than reading the story itself. That's a good trade for a beat this exposed to promotional framing.