I heard of a study where participants were asked to sign up for a fake social media platform with absurd terms (such as giving up your first-born child). I was curious as to what can actually be put into a ToS. For example, can I put in a clause like this:
Upon signing up for this, you agree that I have the right to enter your house at any time without notice and use any motor vehicles you own without notice.
Or something else like that.
This paper describes a website created to test the hypothesis that nobody bothers to read the TOS, and it contained a privacy clause saying that "we may share everything, and a clause that
by agreeing to these Terms of Service, and in exchange for service, all users of this site agree to immediately assign their first-born child to NameDrop, Inc. If the user does not yet have children, this agreement will be enforceable until the year 2050. All individuals assigned to NameDrop automatically become the property of NameDrop, Inc. No exceptions.
Since this was not a real webpage, there was no attempt to enforce.
TOS is just another kind of contract, so the question is which kinds of contract conditions are enforceable and which kinds are not. It depends on whose laws you are operating under, which explains the use of expressions like "void where prohibited by law". There are innumerable conditions that could be included in a contract which are unenforceable. A contract requiring a person to commit a crime is unenforceable. Many jurisdictions have laws prohibiting a person from waiving certain rights, for example you cannot sign away your right to be represented by legal counsel in a dispute. There are related "unfair / deceptive" practices laws, which might include meta-conditions that any clause disclaiming liability must be prominently displayed. These are terms that a reasonable person would clearly know in advance are illegal and unenforceable.
There is also a concept of "unconscionability", a finding that a certain condition favors the business to the point of "shocking the conscience", which may take a deeper legal analysis to evaluate. A requirement to litigate small disputes in California might be deemed unconscionable for a customer in an East Coast state, but it might not be. A typical characterization of an unconscionable contract is one that "leaves one party with no real, meaningful choice and is unreasonably advantageous to the other party", especially when due to asymmetrical negotiation power. For example, Ellis v. McKinnon , 18 Cal. App. 4th 1796 (employment contract for a salesman, commissions were forfeit if company had not received payment from the customer by the termination of employment). The main case law in the US in this area is Williams v. Walker-Thomas Furniture, 350 F.2d 445, where a customer bought furniture on credit, with a clause allowing repossession of all of the furniture in case of default by the customer. The doctrine is encoded in UCC 2-302. However, "unconscionable" does not mean "I don't like it". In lieu of statutory price controls, a customer probably cannot avoid paying an agreed on exorbitant price for a product ($100 for a dozen rolls of toilet paper). Here is a bit more legal analysis of unconscionability, which focuses on three factors: one-sidedness, oppressiveness, and likeliness to result in unfair surprise.
The question is too broad to answer specifically. The answer to "what can actually be put into a ToS" would be pretty much an infinite list. You also need to differentiate between what you can put into a contract (which is what your question appears to ask) and what you can enforce if it is in a contract (which I suspect is the real question).
You can generally put whatever you like into a contract but some rare exceptions do exist. For example, in England and Wales, if a landlord enters into a contract with a tenant which contains a clause which is prohibited by Section 1 of the Tenant Fees Act 2019 and has received a civil penalty for doing so previously, then they will commit an offence if they do it a second time within 5 years. Another example would be if you dishonestly make a false representation with an intent to make a gain or cause someone a loss, which is an offence under sections 1 and 2 of the Fraud Act 2006. For example, a clause which states "Party A requires Party B to do X upon which Party A will do Y" could amount to fraud if Party A knows at the time of offering the contract that they have no intention of ever doing Y.
More generally, clauses may be rendered unenforceable through various mechanisms such as:
Assuming there is no specific rule of law which renders your clause unenforceable, then the general rule is that the courts should not interfere with the agreement that the parties negotiated, even if it seems like a bad bargain for one side.
If you include such terms, the courts will probably not consider them legally binding. Specifically, adding irrelevant terms like that is likely to be ruled, "fraud in the inception" or "fraud in the inducement." The issue is that especially irrelevant terms often violate the "meeting of the minds" requirement for a contract.
As a real-world example of people trying to hide insane stuff in the terms and conditions, consider Toyota's "Your Other You" campaign. Here, Toyota cyberstalked the person (sending creepy text messages, phone calls, and e-mails, making sure to mention the victim's home address to make sure it sounded threatening), then tried to argue it was OK due to it being in the terms, but the judge still allowed the lawsuit to proceed (Toyota Lost the appeal).
Shows like "Scare Tactics" tend to suffer similar legal risk (and intermittently are sued), though the truth is revealed quickly enough that the amount of harm victims suffer is usually lower.
