I'm sure that publishers will try and fight back, just like with music copying.
Open. The journals. Stop. Having them be closed.
Also there are many places to ask for a paper if you really need it like on reddit ( I don't remember the sub reddit name). people do have groups on Facebook and Twitter to ask and provide pdfs. For really desperate people with rare cases if that all does not work, sending a message to thr author asking for a version never gets no. people like when others are interested in their work.
If we're going to allow authors to do that then we could reduce the workload by uploading all the papers to central servers, like some sort of hub, and just letting people download them.
The real problem here is that scientific works are not artistic creations and shouldn't be covered by copyright. Also USA exporting ever longer copyright terms.
Limit what counts as an artistic work (or what counts asa derivative, extracting data is not artistically derivative).
Limit period of copyright to 20y like for patents.
Bosh.
Entirely my own personal opinions.
Whether works funded by government dollars should be copyrightable is of course a different question.
In copyright there is already a concept (in UK caselaw at the very least) that things that can only be presented in a limited way are effectively not artistic enough to gain copyright. A scientific paper can have artistic prose and such, but the purpose is to make a factual presentation. I'd be happy to allow people to chose to state that their paper is not a presentation of facts, but such things should not be accepted into scientific corpus and should be excluded when considering if an author has fulfilled a contract to produce a scientific paper, or fulfilled a duty to do scientific research.
Photographs which are slavish reproductions---which a photo of apparatus needs to be in order to be scientifically useful---are not artistic works, for example. You can choose puce headings with lilac lines for your table but if it's in a scientific paper the purpose is informational [if that wasn't your purpose then you shouldn't have done it]. Coloured diagrams, sure, if they're used outside the context of the paper, and were manually manipulated, then allow them to be considered artistic works; but in the context of the paper they're presentations of supposed facts.
I don't doubt you can lawyer your way into arguing your scientific paper is a trademark, or a registered design, but it seems entirely reasonable to simply prevent such things with ab exclusion clause in the legislation like "scientific papers are not subject to IP laws and can be shared freely". Of course you need to say what a paper is (something published as if it were a paper, or submitted to a journal as it were a paper, or made available publicly as if were a paper).
While they are perhaps not artistic creations, I see scientific papers as the creative work of one or more humans. Most computer code is not an artistic creation, but is a creative work nonetheless. My comment here is not artistic…
Not in general artitistic though. Shoe-horning [ie forcing] computer code into the frame of copyright is also a bad move. It needs its own IPR really.
It's normally standard academic journal licensing terms that the author is permitted to distribute the work in pre-print form on a person-to-person basis.
To my knowledge preprint repositories are popular in physics, math, CS, bio, but not much elsewhere. I work in research in mechanical engineering and no one uses arxiv or anything like it (unfortunately).
That said scihub has not failed me often.
Being legal makes a difference.
It's an minor risk I would imagine, compared to the benefits of reading the articles. But paying for the articles you cite or reference in presentations seems like a good decision for the same reason as you don't want some "pirated version" notification popping up while you give a presentation at a conference. The risk is a lot lower, but it's the same logic.
Considering the cost of that handful of articles versus the budget for a company's coffee supplies it seems like probably the right decision. I don't think I'd go for an actual journal subscription though, that seems like ridiculous overkill unless you're the size of Google.
In many cases, these sorts of crackdowns can come long after-the-fact. The statute-of-limitations is short (3 years, I think), but a lot of courts start the clock ticking when the violation is discovered. If your organization pirates for 30 years, and a publisher does an investigation in 2022, and discovers all 30 years of violations, they would likely file in such a venue.
In most cases, the liability comes in when a business is failing, and goes into don't-give-an-f territory. If Elsevier / Google / Amazon / Coke / [insert random successful business] starts suing researchers when they're big, they're liabile to be hated for it, which endangers their brand and their existing business.
If Elsevier goes into bankruptcy and those copyrights get bought up by parasites, or otherwise starts struggling to survive (as inevitably happens with almost any business eventually) those sorts of retroactive lawsuits often come into play. Most businesses I worked with in the nineties went into owned-by-sleazeballs territory at some point in the forty years since.
That is a hack, not a solution. As you noted, it is illegal.
Workarounds are good as a fast fix, but sooner at later we need to look at the system and fix what is not working. This not only fixes this problem but it may fix other copyright/publication/papers problems.