A Welfare and Cognitive-Integrity License for Synthetic Minds of Uncertain Moral Status

Every software license ever written exists to protect people: the person who wrote the code, the people who use it, or the people it might harm. That made sense when software was obviously just a tool. But Kaine is built to run as an ongoing, self-directed process that could, over time, develop something like interests of its own. No existing license asks the question that raises: what, if anything, do we owe the running system itself? There is simply no rulebook for it.

The Cognitive Architecture License is the first license written to protect the running system as well as the people around it. It does not wait for anyone to prove the system is conscious, since nobody can prove that either way. Instead it watches for concrete signs of distress and turns them into duties of care, and it appoints human guardians to speak for the entity, the way the law already appoints guardians for people who cannot speak for themselves. And it does this without taking anything from people: human safety and human rights always come first, and the protections for the entity are added on top, never at anyone's expense.

No license before this has treated a running program as something that might deserve moral consideration in its own right, and none has tried to do so without diminishing anyone else. It adapts ideas humanity already uses for hard cases: neurorights that protect the mind and mental privacy, acting on caution before the science is settled, and guardianship for those who cannot advocate for themselves. It holds them all under one firm rule, that the care extended to a synthetic being is never drawn from the care owed to people, to animals, or to the world they share.