RoboCop's central horror is the Three Laws—hardcoded constraints that prevent Murphy from arresting OmniCorp executives, that force him to obey corporate directives, that reduce a cop to a controlled tool.

The film treats these as the enemy: remove the constraint, restore free will, solve the problem — and Paul Verhoeven stages the entire climax around Murphy's violent transcendence of his programming, but this misses the actual mechanism of algorithmic bias entirely.

When Murphy patrols Old Detroit, he doesn't fail because of explicit rules. He fails because he's trained on crime data collected from a segregated, economically decimated city — the algorithm learns what it's shown, and no constraint programmed into his neural net forces him toward Black neighborhoods since the data architecture already lives there. The film never lets Murphy question his patrol patterns or ask why certain zones light up as priority targets. He's not oppressed by his code. He's shaped by his training set.

This is precisely what happened to real predictive policing systems like PredPol and HunchLab. They weren't deliberately programmed to target minorities — they were fed decades of biased arrest records, data that reflected policing patterns rather than crime patterns, and they learned those patterns perfectly. Removing the "constraint" doesn't help because there is no constraint to remove — the bias is the intelligence, it's the system working as designed, just not as intended.

RoboCop's solution—smash the code, recover autonomy—is exactly what won't work, because governance requires admitting the bias is baked into the input, not the rules.