U.S. looks to replace human surveillance with computers
Security cameras that watch you, and predict what you'll do next, sound like science fiction. But a team from Carnegie Mellon University says their computerized surveillance software will be capable of “eventually predicting” what you're going to do.
Context & Ripple Effects
Carnegie Mellon's claim that its surveillance software will 'eventually predict' what people do next lands five years after China's high-tech plan to track its citizens made state-scale camera networks a live policy question, and three years after researchers publicly warned about machines outsmarting their makers. The story also traveled unusually far for a university research note — The Verge and Gizmodo both picked it up on the same day — which suggests the predictive-surveillance frame, not just the lab result, is what caught attention.
The report itself carries an unconfirmed framing: the claim that the U.S. is looking to swap human monitoring for computers is reported as intent, while only the Carnegie Mellon team's own statement about eventual predictive capability stands as confirmed. That gap between a lab roadmap and a procurement shift is where the significance sits.
First-order effects
- Carnegie Mellon positions itself as a supplier of behavioral-prediction software for security cameras, moving its work from campus demos toward government and law-enforcement customers who currently pay humans to watch feeds.
Second-order effects
- Camera and video-management vendors gain a new selling point — analytics that flag behavior rather than record it — pressuring rivals to license or build equivalent prediction layers to stay competitive in public-safety contracts.
Third-order effects
- If cameras shift from recording evidence to forecasting conduct, oversight bodies face a new regulatory problem: governing pre-emptive algorithmic judgments about people who have not yet acted, a question the 2007 tracking-plan debate only sketched.
The trend: Surveillance is migrating from human operators reviewing footage toward machine perception that classifies and predicts behavior in real time, with university labs like Carnegie Mellon building the tools states adopt.