Movies have ratings. G. PG. R. NC-17.
Video games have ratings. E. T. M. AO.
Both were born because the public got fed up before the government got moving.
We’re in that exact moment with AI content. Right now.
Scroll Facebook for 5 minutes. AI-generated content is everywhere — but the label? It’s buried so deep most people never even see it. That’s not transparency. That’s a checkbox.
And even when you DO find it, it tells you nothing. One generic “AI” tag slapped on everything from a grammar-polished email to a fully synthetic deepfake.
That’s like rating every movie the same regardless of whether it’s Bambi or Saw.
Here’s what I think we actually need — a spectrum, not a scorecard:
AI-1 — Human-Authored. AI only used to polish or format.
AI-2 — Human-Led. AI contributed, but a human directed and reviewed it.
AI-3 — AI-Generated. Mostly machine output. Some human oversight.
AI-4 — Fully Synthetic. 100% AI. No meaningful human in the loop.
AI-∞ — Recursive. AI generating content that feeds more AI. No human. Ever.
This isn’t about labeling content good or bad.
It’s about knowing what you’re consuming — the same way a nutrition label doesn’t tell you what to eat, it just tells you what’s in it.
AI-4 isn’t inherently bad. A Marvel VFX shot is AI-4. So is a fake news article. The label tells you the origin. You decide what to do with it.
The EU mandates AI labeling by August 2026. California follows.
The question isn’t IF we get a rating system.
It’s whether the industry builds a smart one — or the government builds a bad one.
Which do you think happens first?
