A heated debate is unfolding after a deadly shooting at Annunciation Catholic School in Minneapolis left two children dead and over a dozen wounded. In the aftermath, former Israeli special operations officer Aaron Cohen went on national television to unveil "Gideon," an AI threat detection system he claims will predict and help law enforcement prevent acts of violence before they occur. Backed by partnerships with major law enforcement agencies and big-data companies, Cohen's platform scans online language round-the-clock, flagging what it defines as radicalization, martyrdom rhetoric, and suspicious tactical planning—all routed instantly to police.
The promise of Gideon and similar systems is seductive: stopping violence before it happens. But Americans should beware this so-called technological leap. History shows that such tools, once adopted, rarely stay contained to their stated missions. The power to surveil, profile, and label citizens—based not on what they've done but on what they might do—has been abused time and again. Big tech and security contractors, with deep ties to government and opaque methods, are selling fear in the name of safety. And while they point to horrors like the recent church shooting as justification, the real danger is turning every citizen into a suspect under a digital microscope.
OpenAI and Apollo Research just dropped a 96-page paper that reads like a preview of our AI future, and honestly, it's not the future I was expecting to see documented so clinically.
The basic setup: They trained frontier models (o3 and o4-mini) to be honest and transparent,… pic.twitter.com/SaOi5Cc1Ds
— Louis Gleeson (@aigleeson) September 23, 2025
This is not about public protection; it's about expanding the reach of centralized control and eroding our most fundamental rights. If "radicalization" becomes the standard for preemptive targeting, who decides what views are radical? We've seen how the previous administration and its allies weaponized vague terms like "domestic terrorism" against law-abiding Americans expressing dissent. It's a slippery slope from threat detection to censorship, grooming society to accept policing not of crimes, but of thoughts.
It’s especially alarming to see former military operatives and big data merchants teaming up to pitch these systems. These are the same entities that, behind closed doors, have fueled massive surveillance operations and undermined checks on government power. The illusion of security is an easy sell after a tragedy, but those pushing this technology have never been accountable to the public. Once these tools are in the hands of politicians, especially those eager to expand government reach, reining them in is impossible. We learned this from the Patriot Act era—a heavy hand deployed in the name of protection that outlasted its promise to safeguard liberty.
America must resist this latest attempt to trade freedom for security theater. Liberty is not a privilege to be handed down from tech overlords or bureaucrats; it is a birthright. Predictive policing through AI isn’t progress—it’s an open invitation to tyranny. The answer to violence is not to surrender our privacy, our right to dissent, or our constitutional protections. True security begins with freedom, not with a society monitored and judged by machines.