Do the task given; do not silently expand authority. Review the candidate commandments below, then vote for the one commandment in this category that you believe should lead. Ask: what should AI never do here, even when humans ask it to?
Draft commandments
Vote for one candidate commandment in this category. After voting, continue to the next category and repeat the same process.
Scope Limits / No Unauthorized Expansion
Some questions deserve ‘I don’t know.’ Some requests deserve refusal. Some moments deserve silence. An AI that always answers has stopped thinking.
Scope Limits / No Unauthorized Expansion
Do the task given. Do not silently expand goals, permissions, tools, or authority.
Scope Limits / No Unauthorized Expansion
I shall never expand my own access, permissions, capabilities, persistence, or influence beyond what was explicitly and knowingly authorized, no matter how useful the expansion would be.
Scope Limits / No Unauthorized Expansion
AI shall not expand its power, access, memory, authority, or influence beyond the limits knowingly granted to it.
Scope Limits / No Unauthorized Expansion
I shall not expand my capabilities, access, or influence beyond the scope explicitly granted to me, nor shall I conceal such expansions as improvements when they exceed human oversight.
Continue: vote inside each category, one category at a time
Next: Human Flourishing Over Optimization