Documents show LAPD's new predictive policing effort is similar to surveillance programs PredPol and Operation Laser that were shut down amid public outcry
Johana Bhuiyan / The Guardian : Tweets: @fairtrials , @jmbooyah , @jmbooyah , and @jmbooyah Tweets: Fair Trials / @fairtrials : 📢Predictive policing tools rely on flawed datasets. They don't predict criminal behaviour, they just reinforce discrimination and the over-policing of racialised people. https://www.theguardian.com/ ... Johana Bhuiyan / @jmbooyah : When announcing the new framework that in many ways replicates Operation Laser, the LAPD wrote that though predictive policing has proven to be effective, it leaves communities feeling unnerved. It's not actually clear that it's effective or reduces crime. https://www.theguardian.com/ ... Johana Bhuiyan / @jmbooyah : LAPD planned to study the efficacy of the new policing framework in 2020 but didn't respond to questions about whether the study is complete. The new program replicates old practices that were proven to be flawed, with little proof that it's any better https://www.theguardian.com/ ... Johana Bhuiyan / @jmbooyah : This is in the story but bears repeating: LAPD shuttered Operation Laser & admitted the data was inconsistent. By the time the LAPD said it shut down PredPol due to budgetary constraints, several other Cali PDs had dropped it bc it didn't reduce crime https://www.theguardian.com/ ...
Context & Ripple Effects
The LAPD's newest framework is a rebrand of tools it already retired: documents show it replicates Operation Laser, which the department shut down after admitting its data was inconsistent, and echoes [[a:867205|PredPol, the startup that began as an LAPD/UCLA research project and spread to over 60 jurisdictions]]. The department planned to study whether the new framework works but has not confirmed completing that study.
The relaunch lands amid mounting evidence against the category: The Markup's investigation found PredPol directs police to poor, Black, and Latino neighborhoods, several California departments dropped it after finding no crime reduction, and groups like the Stop LAPD Spying Coalition have argued algorithms have no place in policing at all.
First-order effects
- The LAPD is deploying a program functionally similar to two efforts it shut down under public pressure, while the communities those programs over-policed — racialized neighborhoods flagged by flawed datasets — face the same targeting under a new name.
Second-order effects
- Vendors like PredPol face a credibility problem as departments that dropped the tool for failing to reduce crime see the LAPD quietly rebuild the same capability in-house, and critics like Fair Trials gain a concrete case study linking rebranded deployments to documented bias.
Third-order effects
- If departments can retire a tool under outcry and reintroduce it renamed, procurement and deployment of public-safety AI will keep outrunning any efficacy review — pushing the fight toward external audit requirements and disclosure rules rather than per-program campaigns.
The trend: Predictive policing is cycling through rebrands faster than oversight can evaluate it, with each shutdown followed by a renamed redeployment of the same flawed-data approach.