Exactly. There never was a declaration of independence of cyberspace. BUT government and law moved too slowly by years and years. And they have, of course, not learned their lesson.

For example: suing Nappster 2 years after it launched. And that was just because it was an extremely clear-cut case. By the time they did that there were 10 such networks, none of which were sued, none of which had clear laws or court decisions stating clearly one way or the other if it was legal.

And when we're talking a vague issue, for example how copyright affects search engines, the first actually settled case (which was still a far cry from establishing the rules) happened in 2006, 16 years after the initial search engine started operating and over 8 years after Google started it's meteoric rise. The specific decision the courts deigned to make, after 16 years? That caching a page so it can be used to build a search index in the first place does not by itself violate copyright. Great, well, that covers it then. My point is, by then the cat was out of the bag, ran to the neighbors house, got 6 kittens, who each got 6 kittens themselves and one of it's grandchildren ate the sandwich the judge was hoping to have for lunch and one of the other kittens got adopted by the president of the US, while the rest invaded and destroyed the houses of publishers that tried to protect their copyright.

Imagine the insanity, the damage that any real court decision against search engines would do today. "No you can't show previews". "Ads don't respect trademarks". There is no room for any such decisions now. The few decisions they have made (in >30 years) have amplified the damage to the victims that the court system tried to help (just ask a few newspapers).

Of course, none of this has instilled any sense of reasonableness, modesty or urgency in any parliament, court or even executive around the globe. For instance, they could PRE-clarify the laws before AI takes over 5 industries. Does AI training violate copyright? What are the rights of an employee that gets fired because AI does their job? No government felt the need to answer the copyright question when it mattered, 7 years ago, and there is ZERO action on the second question. Are they planning to answer the people displacement question once 99% of companies have done it because competition forced them to?

Now any answer they give on the copyright front is beside the point since no court or Parliament actually has the power to order existing (potentially law-violating) models to be destroyed. Once again, they have placed themselves into a position where they are totally irrelevant. Now one might ask, the time is to decide if you violate copyright by training a model using a model that was trained while violating copyright. Perhaps that one is still relevant. But nothing will be done.

And please, it doesn't matter what your position is on the issue. Can model training violate copyright? Yes or no? We live in a democracy and no decision is made. This is an important part of why big companies get to openly violate laws on an unprecedented scale for billions and billions without consequences while kids sometimes get locked up for stealing a single candy.