May 6, 2026

Everytown’s Own Disclaimer Undercuts Its AI-Driven Gun Data Tool

By Larry Keane

Gun control activists have always had a numbers problem. Everytown for Gun Safety just added an artificial intelligence (AI) problem as well.

Everytown, the billionaire Michael Bloomberg-backed gun control organization, launched EveryShot as an AI-powered database that tracks news reports involving criminal firearm misuse or the brandishing of firearms. The group presents its new tool as a “resource” for policymakers, journalists and researchers. But Everytown’s own disclaimer warns users that the tool’s outputs may be incomplete, inaccurate or otherwise unreliable.

That’s a glaring warning sign that also goes directly to credibility. Everytown cannot ask lawmakers to treat its gun control claims as authoritative while telling users not to rely on its own AI-generated data as factual or accurate evidence to back up its policy prescriptions.

Everytown wants lawmakers, reporters and the public to trust its gun control statistics. Its newest artificial intelligence tool tells them they should not.

Garbage In, Garbage Out

EveryShot relies heavily on media reports, meaning the tool is only as reliable as the coverage it scrapes.

Early crime reporting is often incomplete. Initial stories may omit whether a suspect was prohibited from possessing a firearm, whether charges were later changed or whether the firearm description was confirmed by law enforcement. Reporters frequently rely on preliminary information that can easily and predictably shift as investigations move forward.

EveryShot then uses artificial intelligence to classify those reports and produce searchable data, which is not the same as verified evidence.

Artificial intelligence can process flawed information faster. It cannot make incomplete reporting accurate. It cannot distinguish advocacy assumptions from official findings unless the underlying record allows it. It cannot turn a media clip into a court file, police report or Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF) trace record.

That matters because gun control policy is too often built on slogans first and facts later.

Mislabeling Criminal Misuse

Ammoland’s review of EveryShot’s “Machine Gun” and “Legal” ownership filters demonstrated why this deserves scrutiny. The tool reportedly returned several incidents that, after review, appeared to involve illegal machinegun conversion devices (MCDs), prohibited possessors or other facts that did not support the impression of lawfully owned machine guns.

Those are not small distinctions.

Machine guns are already highly regulated and tightly restricted under federal law. Without permission from the ATF, 3D printing, making, importing and/possessing MCDs are all already illegal. Installing an MCD on a semiautomatic pistol is already illegal. A prohibited person in possession of a firearm is already committing a crime. None of that supports new restrictions on lawful gun owners, firearm retailers, manufacturers, ranges or Second Amendment rights.

But that is exactly how flawed gun control statistics, such as those commonly put forward by Everytown and similar gun control groups, are used.

The label becomes the headline. The headline becomes the talking point. The talking point becomes the bill. When the bill becomes law, it serves as the basis for suing members of the industry. Then law-abiding Americans are told to surrender more of their rights because an advocacy group’s “data tool” blurred the difference between criminal misuse and lawful conduct.

That is not public safety. That is policy malpractice.

Everytown Is Advocacy, Not Neutral Research

Everytown is not a neutral public data agency. It is a billionaire class-funded, aggressively antigun political organization with a long record of pushing gun bans, restrictions, lawfare and regulatory pressure against the lawful firearm industry.

EveryShot is not being released into a vacuum. It comes from an organization that already has predetermined policy goals. Its categories, assumptions and omissions should be evaluated accordingly.

A database built by a gun control group to support gun control arguments should never be treated as neutral evidence simply because artificial intelligence is involved.

The firearm industry has seen this before. Expansive definitions inflate numbers. Selective categories drive narratives. Media repetition gives advocacy statistics the appearance of authority. Lawmakers then cite those statistics to justify restrictions that fall hardest on people and businesses that followed the law.

EveryShot risks automating that same cycle.

It does not appear designed to give equal weight to lawful defensive firearm use, existing criminal penalties or the already-extensive regulation governing firearm manufacturers, distributors and retailers. It appears designed to collect the kind of incidents Everytown wants to highlight, through the lens Everytown wants policymakers to see.

That is advocacy software.

Policymakers should not use AI-generated claims from a gun control organization as a foundation for restricting constitutional rights.

Any statistic used to justify firearm restrictions should be transparent, independently verifiable and precise. It should distinguish between legal ownership and criminal misuse. It should identify whether the person involved was prohibited, whether the firearm was lawfully possessed and whether the firearm type was confirmed by official records.

EveryShot does not meet that standard on its own and their own warning says so. Every claim still needs human scrutiny, official verification and context.

Artificial intelligence does not remove bias. It scales it. It does not cure bad definitions. It can harden them. It does not make advocacy neutral. It can make advocacy look official.

EveryShot should be treated as a caution sign, not a source of lawmaking authority. When Second Amendment rights and the lawful firearm industry are at stake, lawmakers should rely on verified facts and official records, not an antigun group’s AI tool carrying a disclaimer that tells users it may be wrong.

You may also be interested in:

OCC Gives Debanked Firearm Businesses a New Path to Hold Banks Accountable

Senate Confirms ATF Director Cekada, Signaling New Era of Fair Enforcement and 2A Respect