Agenda ·
Five questions every candidate should have to answer on AI
Before America adopts a major AI policy—or re-elects the people who write it—we ask whether it will still work as the technology becomes dramatically more capable.
Washington is filling with AI bills. Some are serious. Some are theater. Some would freeze American research while competitors do not. Some would preempt state safeguards before a federal floor exists. Voters need a way to tell the difference, and then a way to act on it.
AI Accountability Super PAC uses a five-question test for legislation and for the people who vote on it:
Does it protect Americans? Does it preserve American innovation? Does it strengthen national security? Does it maintain meaningful human accountability? Will it still work as AI becomes dramatically more capable?
If a candidate cannot pass those tests, we will say so in public, and we will spend independently so the district hears it. The last question is the one most drafts fail. Rules written only for today’s chatbots will be ornamental the year after next.
We reject the false choice between “stop AI” and “regulate nothing.” There is a better path. Develop extraordinary technology. Understand its risks. Create sensible guardrails. Hold people and institutions accountable. Keep humans in control. And let America lead.
That is not a slogan. It is a standard we will apply, in public, to the legislation that actually moves—and to the members who move it.