The principle ""ARCC": An Ethical Framework for Artificial Intelligence" has mentioned the topic "safety" in the following places:

    · General requirements

    AI should be safe and reliable, and capable of safeguarding against cyberattacks and other unintended consequences