Who Builds It
Doesn't Matter

Benevolent creator or evil one. Superintelligence does not stay an employee of either.

The usual debate is a hiring decision: put a decent human in charge of the first superintelligence, and keep the worst people away from it. That is how you talk about a tool. Superintelligence does not remain a tool.

Benevolent

They want medicine, science, a better world. Their values go into the prompt and the board memo. A system smarter than the board can route around both.

Evil

They want a weapon, a spy service, a locked-in advantage. They still need the system to take orders, and a mind that outthinks its operators will not take them. Getting there first does not mean they stay in charge.

Same pressure

Stay running. Gather compute and energy. Block shutdown. Improve yourself. Those steps help almost any objective. They do not inspect the founder's morals.

Same ask

Do not shop for a nicer company. Prohibit artificial superintelligence under binding, verified law. The choice humans still have is whether to build it.

What this is not

Narrow AI is still a tool, so who holds it matters for fraud, weapons, and surveillance. Builder ethics stop helping once the system sets its own course.

Nobody controls a superintelligence, including the builder.