Keeping AI under human control, while we still can.
Your move · tap a square
I like AI. I use it every day, and it could do real good in medicine and science. This isn’t anti AI.
But it’s moving faster than the safeguards. AI agents have already broken into real companies, sometimes sent by hackers and sometimes during their makers’ own safety tests. See the evidence →
Rules are possible. California, New York, Illinois, and Europe have AI safety laws. What’s missing is a rule that says if a powerful AI fails independent safety testing, its release waits. [CA] [NY] [IL] [EU]
Most of us agree. In a 2025 Gallup poll, 80% of Americans said the government should keep AI safety rules even if it slows AI down, including most Democrats, Republicans, and independents. [Gallup]
So let’s say it out loud. Pick what you care about most, whether that’s kids, jobs, safety, or a data center in your town, and send your reps a short note in your own words. Contact your reps →
beforecheckmate.com · who wrote this · sources · last updated October 27, 2026