The principle "The Montreal Declaration for a Responsible Development of Artificial Intelligence" has mentioned the topic "controllability" in the following places:

6. Democracy

How should AI research and its applications, at the institutional level, be controlled?