I may be mis recalling, but I thought the main point of the I, Robot series was that regardless the law, incomplete information can still end up getting someone killed.

In all the cases of killing, the robots were innocent. It was either a human that tricked the robot or didn't tell the robot what they were doing.

For example, a lady killed her husband by asking a robot to detach his arm and give it to here. Once she got it, she beat the husband to death and the robot didn't have the capability to stop her (since it gave her it's arm). That caused the robot to effectively self-destruct.

Giskard, I believe, was the only one that killed people. He ultimately ended up self-destructing as a result (the fate of robots that violate the laws).

That's certainly not the plot of Little Lost Robot.

Little lost robot was about a robot with the first law modified. That's not about the law failing but rather failing to install the full law.

The story from iRobot is one of Asimov s stories and it works exactly as intended. The AI figured that to keep humans safe you have to put them in cages. Humans will always fight over something