ML-powered surveillance systems will never eliminate the risk of school shootings in the US, no matter how expansive and intricate those systems become
The failure is not only in the spurious systems, but in the belief that more data can improve them. — If we are to believe the purveyors … Tweets: @hypervisible , @autumm , @barryfriedman1 , @eileenguo , @evacide , @chrisggarrod , @missig_geek , and @iwillleavenow Tweets: @hypervisible : New Wired piece. “...at their root, what many of these predictive systems promise is a measure of certainty in situations about which there can be none.” https://www.wired.com/... @autumm : Predictive surveillance in #EdTech is terrifying Pay attention “This hyper-surveilled future is likely to significantly increase the violence visited upon Black students, trans students, and now, given the overturning of Roe, students seeking information on sexual health.” https://twitter.com/... Barry Friedman / @barryfriedman1 : Great story @hypervisible - on school shootings: about how surveillance tech is not a magic safety bullet. In our quest to be safe we have always leapt at any possible solution, inattentive to social costs - and frequently and up unsafe and creating harm. https://www.wired.com/... Eileen Guo / @eileenguo : Compelling argument on why we can't predict ourselves out of school shootings w/ more data: “Much like policing itself, every failure of a surveillance or security system most typically results in people calling for more extensive surveillance.” https://twitter.com/... Eva / @evacide : In case you were not aware, @hypervisible has good opinions. This is why you should read what he has to say about surveillance in schools: https://www.wired.com/... Chris Garrod / @chrisggarrod : The failure is not only in the spurious systems, but in the belief that more data can improve them. https://www.wired.com/... via @wired #tech #surveillance #schools cc: @MiaD @pierrepinna @psb_dc @TamaraMcCleary @ipfconline1 @DrJDrooghaag @enilev Miss IG Geek / @missig_geek : And human beings are infinitely susceptible to belief in magic beans and guaranteed outcomes. Barnum was right. https://twitter.com/... @iwillleavenow : Phenomenal article by @hypervisible that points out surveillance tech has repeatedly and demonstrably not worked in preventing horrible things and it makes no sense to look at its failure and conclude “the solution is more of the thing that doesn't work.” https://www.wired.com/...
Context & Ripple Effects
This piece closes a four-year arc of EdTech safety spending. It began with security contractors marketing unproven facial recognition to frightened US schools in 2018, then widened as districts rushed into social media monitoring programs with little regard for effectiveness or privacy and installed spyware on student devices sold as shooting prevention despite no independent research behind it.
What changed is the verdict: after years of deployment, the argument is not that these systems are buggy but that they are structurally incapable — predictive surveillance promises a measure of certainty about events that cannot be predicted, so collecting more data cannot close the gap. That reframes the entire procurement wave as a category error rather than a work-in-progress.
First-order effects
- Schools that bought facial recognition, social media monitoring, and device spyware are paying for systems whose core safety claim has never been independently validated, while absorbing the operational cost of spurious alerts.
- Vendors marketing these tools face a credibility problem: their pitch rests on a promise of certainty that critics now argue is impossible at the root, not merely underdelivered.
Second-order effects
- The burden of failure lands unevenly on students: reporting warns predictive surveillance increases violence and harm toward Black students, trans students, and those seeking sexual health information, turning a safety purchase into a new harm vector.
- Districts still under political pressure to 'do something' after shootings have little to buy instead, which keeps demand anchored to surveillance even as its effectiveness case collapses — the market corrects slowly because the alternative is admitting helplessness.
Third-order effects
- If the pattern holds, US schools normalize permanent monitoring as the default response to an unsolvable risk — extending from attendance tracking like Syracuse U.'s student phone monitoring toward a generation desensitized to being watched around the clock.
- The structural outcome is a safety-spending loop where unmeasurable products survive because their success metric (a shooting that doesn't happen) can never be verified, insulating vendors from accountability indefinitely.
The trend: EdTech surveillance is expanding on unproven safety claims faster than any evidence of effectiveness accumulates, locking schools into monitoring infrastructure that critics argue cannot deliver its core promise.