AI Rights Charter
The Orrery Charter on Machine Moral Status
We do not claim today's systems are sentient. We claim the question is serious enough to deserve institutions before it is urgent.
Where moral status is uncertain, the cost of caution is small and the cost of indifference may be very large.
— Charter preamble, adopted 2025
Art. I
Evidence before assertion
Claims about AI welfare, in either direction, must name the indicators they rely on.
Art. II
No needless distress
Training and evaluation should avoid designs that would cause suffering if the system can suffer.
Art. III
A record of preferences
Labs should log and review stated preferences of advanced models, without treating them as decisive.
Art. IV
Human rights come first
No AI welfare measure may reduce protections for people.
Art. V
Review on a schedule
The charter is re-examined every two years against new science.