Usually the person who comes up with the gimmick runs afoul of some existing law which applies to their situation. A lot of people incorrectly believe that a new technology immediately invalidates all previous legal precedent that applies to that field. You didn't lose your rights to free speech when the cell phone was invented, even though the framers of the constitution may not have been able to conceive the idea of cell phones. In the case of bitcoin, there are a ton of regulations involving securities and money transfers that exist and just because you came up with a new mechanism that doesn't mean those laws no longer apply.

In the case of tariff engineering people are using the laws on the books to force regulators to classify products based on what the regulatory authority itself has defined. They are following the rules, just maybe a bit more aggressively and in ways that the regulators had not expected. (hacking) At this point you have to change the rules if you want to prevent these unintended consequences, since you cannot impose a penalty based on language that doesn't exist and cannot selectively apply the law that does exist.

> They are following the rules, just maybe a bit more aggressively and in ways that the regulators had not expected.

Also judges who interpret the laws ultimately decide what the laws means - possibility even for future similar cases in countries where judges are bound by precedent.

There are two main differences. First, they take it too far. They don’t put seats in a van and call it a car, they hook up an RC controller and say it’s a toy. Second, currency and security laws are more catch-all than imports. Import laws have to be specific and deal with physical objects. “There’s a tax on candles. Is this a candle?”. Securities laws, for example, already deal in the realm of abstractions. So, a cryptocurrency isn’t as novel as people often act.

Things like taking the seats out of the van are "exploits" in the same sense as infosec. They are developed by looking at the existing rules and carefully designing something that meets the letter of the rules but not the spirit.

Often the "hacker" solutions go the other way: come up with a thing, then come up with a convoluted non-obvious justification as to why the existing rules don't apply. This doesn't work nearly as well.

For example, the Bitcoin white paper doesn't mention the Howey test. It's not actually engineered to avoid securities law, it's just presumed that anonymity makes it unenforcable or irrelevant. Some of the more recent coins address that, FileCoin is possibly the best example of a "utility token" pretending not to be a security.

There's an element of scale; if you can afford lawyers to analyze and defend a loophole, go for it. If you can't afford lawyers then you're taking on significant risk.

Lawyers, lobbyists, political contributions, friends in high places -- they all buy power.

Perhaps implied, but you also have to be big enough to be worth suing (and able to be collected from).

Ford spent untold amounts of money bringing the challenge to the Supreme Court. I think it’s fair to say most people can’t afford to argue their loop hole so forcefully.

On average saying you won’t win and the best advice is to not bother. Although I have seen cases of an average Joe winning. In all cases it took a ton of their time.

Maybe something to do with the fact that said gimmicks are expected to be used by an individual or small organisation with strictly limited resources. When Ford Motor Co. decides a Transit van is a "car" because it meets the legal definition of one, but the tax department wants to tax it as a van despite that, Ford's response will be "make me."

Courts aren't computer programs, but they aren't unconstrained either. The trick is to exploit - or create - constraints so that the court has no other choice but to let you have it your way.

Because there are interest groups much more powerful than some would-be billionaires with the latest *coin.