Second Law of Robotics

The Second Law of Robotics is the rule that a robot must obey human orders except when those orders conflict with the First Law. It sits between the First Law’s prohibition on harming humans and the Third Law’s protection of a robot’s own existence, defining both the authority humans have over robots and the boundary that limits that authority.

Definition and Place Among the Laws

The Second Law is presented as part of the standardized theoretical foundation for positronic brains. Dr. Gerrigel explains that designing even a robot brain with comparatively minor innovations demands extensive work, drawing on an established basic theory founded on the Three Laws of Robotics. Within that framework, the Second Law makes obedience conditional: a human order is binding only when carrying it out would not violate the First Law.

The rule therefore does not grant human commands absolute priority. It establishes obedience as a central robot obligation while placing the protection of humans above it. The Third Law, by contrast, permits a robot to protect its own existence only when doing so does not conflict with either of the first two laws. Together, the stated hierarchy locates the Second Law between human safety and a robot’s self-preservation.

The Rule as a Test of Identity

The law becomes relevant to the investigation when Baley argues that a robot would have to obey a human command. Dr. Gerrigel qualifies that claim: obedience does not apply when it would break the First Law. This qualification redirects the question from a simple opposition between command and refusal to the conditions under which a robot may act.

That distinction matters because Baley’s case turns on whether R. Daneel is human or a robot. Dr. Gerrigel describes practical ways of distinguishing a highly humanoid robot from a human, while Baley invokes the Second Law in challenging the idea that a robot would draw a blaster. The rule supplies a framework for interpreting conduct, but applying it requires deciding whether the action would violate the First Law and whether the person in question is a robot at all.

Application and Limits

Baley raises the specific claim that Daneel drew a blaster on an unarmed group of men and women and threatened to shoot. K. Daneel replies that he did not shoot. Dr. Gerrigel’s answer returns the discussion to the hierarchy of rules: a robot need not obey when obedience would involve breaking the First Law. The exchange thus distinguishes threatening an action from carrying it out and makes the First Law’s constraint essential to any judgment about what the Second Law required.

The Second Law is consequently a governing principle, not a self-interpreting verdict about a particular incident. The exchange does not resolve every question raised by Baley’s accusation; instead, it shows why the command, the possible harm, and the robot’s status must be considered together. Its narrative importance lies in making the apparent simplicity of robot obedience conditional on the more fundamental prohibition against harming humans.