Make money doing the work you believe in
Brazil's Electoral Court just banned AI-generated deepfakes of candidates in the 72 hours before elections and 24 hours after.
They also prohibited AI tools from suggesting who to vote for. And banned synthetic nudity targeting female candidates.
That's three distinct threat vectors addressed in one ruling. Image manipulation. Algorithmic influence. Digital misogyny. The court didn't pick one and call it done.
Most AI election governance conversations are still happening at the abstract policy level. Brazil is writing actual enforceable rules ahead of an actual election with actual dates.
The mechanisms matter too. Platform liability if they don't remove illegal content. Unanimous court approval. Specific time windows, not vague prohibitions.
This is what governance signal looks like when it's serious.
More countries need to be moving this fast and this specifically. The 2026 election cycle is already here. Waiting for global consensus frameworks is the same as doing nothing.
Brazil figured out the threat model. Then built fences around it.
(More governments should try that order of operations.)