Cody Wilson releases ‘Hochulization’ tool to bypass 3D-printed gun detection
The creator of the first 3D-printed firearm has unveiled an open-source workaround to New York’s new printer mandates, arguing that hash-based detection is easily defeated by digital noise.
Cody Wilson, the maker of the world’s first 3D-printed gun, has released a tool called “Hochulization” designed to subvert government-mandated software intended to block 3D printers from producing firearms. The tool targets the unique hash of gun design files, altering them to evade detection systems used in New York State, where Governor Kathy Hochul recently passed the nation’s first law requiring such software on new printers. Wilson, based in Austin, Texas, named the tool as a direct reference to the New York governor who pushed the legislation through as part of the state’s FY 2027 budget.
The mechanism is straightforward: Hochulization adds a series of bytes to each file, changing its hash without altering the function of the final printed product. Wilson plans to apply this treatment to all files on his repository, DEFCAD, which hosts over 200,000 gun designs and components. By scrambling the hashes, Wilson aims to prevent the creation of a complete library of known gun files, a strategy he describes as a “self-defense measure” against what he views as invasive digital surveillance. He intends to make the tool open source, allowing individual users to apply the workaround independently.
Wilson’s team claims to have found ways to bypass more sophisticated AI-based detection models using “adversarial noise.” This involves techniques such as re-saving parts as different meshes or adding unused geometry to the file. In internal testing of an AI detection model, Wilson’s team reported a 30 per cent false positive rate, a figure he argues highlights the inherent flaws in prediction-based systems. The development suggests that while hash-matching is easily defeated, more complex AI scanning may struggle to distinguish between legitimate files and firearms without incorrectly flagging non-gun items.
The move comes as 3D-printed firearms gain traction in the United States. The New York Police Department confiscated just one 3D-printed gun in 2021, a number that jumped to 109 in 2024. Prominent gun policy group Everytown found similar upticks across 20 other cities. While 3D-printed models remain a small fraction of conventional firearms, high-profile cases, such as the alleged use of a partially 3D-printed pistol by Luigi Mangione, have intensified calls for stricter regulation.
Regulators are now faced with a technical dilemma. A purely hash-matching approach limits false positives but is trivially easy to bypass, as Wilson’s tool demonstrates. More expansive AI approaches may catch more guns but risk introducing significant false positives, a concern shared by makers and digital rights experts who fear the technology could limit creative freedom. A similar bill, AB 2047, is advancing in California, which would require new printers to have print-blocking software and ban non-compliant models from sale starting in December 2029.
Wilson, a leading voice in arguing that gun file code is constitutionally protected speech, hopes his tool will influence these policy discussions. By showing how simple it is to complicate a hashing-based approach, he is attempting to force policymakers to choose between a more aggressive scanning scheme or the status quo. As the debate continues, the “Hochulization” tool marks the opening volley in an evolving battle between regulators seeking to close gaps in gun control and advocates resisting mandatory file scanning.
