Other people have already pointed out the differences between implementing a check for a specific banned print and a vague categorical ban. It would be like if printer manufacturers weren't just asked to prevent the printing of US dollars, but anything that looks like money, having an ability to detect if something is money-like based on look and feel alone, without relying on an existing database or hardcoded watermarks.

Your implication makes me think that you assume that this useful-yet-not-overreaching detection tech is possible. Do you have any ideas for how this would be implemented? Because in my mind, the only way to ensure compliance would be either a manual check (uplink to the manufacturer or relevant government authority, where an employee or a model trained on known gun models tries to estimate the probability of a print being part of a gun) or a deterministic algorithm that makes blanket bans on anything remotely gun-like (pipe-like parts, parts where any mechanical action is similar to anything that could be in a gun). These scenarios seem to be both a lot more annoying and a lot more invasive. There's no negative consequences for tuning detection to always err on the side of caution and flood the user with false-positive refusals to print. Both scenarios are obviously a lot more involved and complicated than a basic algorithm checking if you're trying to print an image of a US dollar. Therefore I don't see a reason why drawing this comparison is useful. The only thing these implementations have in common is that they're detecting something.

>Other people have already pointed out the differences between implementing a check for a specific banned print and a vague categorical ban.

If you have seen that other people have pointed it out, you have already seen my response, but I guess people keep repeating the question, so I need to repeat the answer. This regulation establishes a working group to investigate this technology. If the technical aspects are as difficult as you claim, the proposed regulation will basically be voided. Your concerns are already factored into the proposal and therefore aren't a valid argument against the proposal.

That said, the regulation also makes it sound like "implementing a check for a specific banned print" would be an acceptable outcome of this law. From page 11 of the actual proposal:

>(b) be authorized to create and maintain a library of firearms blue- print files and illegal firearm parts blueprint files, and maintain and update the library, including by adding new files that enable the three- dimensional printing of firearms or illegal firearm parts. In further- ance of this authorization, the division may designate another govern- ment agency or an academic or research institution in this state to assist with the creation and maintenance of the file library. The library shall be made available to three-dimensional printer manufactur- ers, vendors with demonstrated expertise in software development, or experts in computational design or public safety, for the development or improvement of blocking technology and firearm blueprint detection algo- rithms. The division shall establish safeguards to prevent unauthorized access to and misuse of the library and shall prohibit all persons who are granted access to the library from misusing, selling, disseminating, or otherwise publishing its contents.

Think of it like the early stages of internet copyright protections, the first step is just cross-referencing the design with a list of known banned designs. Just like an early Youtuber could have mirrored banned videos to bypass copyright detection, people will likely still be able to manipulate designs in certain ways to get past this sort of ban. That's ok. Regulation like this doesn't have to be 100% effective to still be worth doing. The goal here is to make it more difficult for some random person with no expertise to buy a 3d printer, download some files, and print a weapon.

I'm willing to admit that it's entirely possible that a full on-demand analysis of whether a shape could potentially be part of a gun might not currently be possible and it might be years before that becomes feasible, but until then, simply banning a handful of the most popular STL files would still have value.