This might be true, at least in specific cases. I had a patent examiner who looked at one of my metal tubes that has multiple holes in it, and said that it wouldn’t work because there were so many holes that the tube would become “overfull of air” and not accept any more. I argued that air flow into and out of all the holes would always equilibrate at some balancing between the holes, and that the tube would always contain the same amount of air. We went back and forth in a most frustrating fashion for months, about a bizarre point that would never have occurred to me as a possible problem. I eventually got the patent, and we did sell these devices for a while, and they’re referred to in various scholarly publications and obviously worked. It is glib to say that the examiners were “idiots”, but it is often true that the problem is the examiners not understanding inventions. They get a huge variety of technical input covering a vast range of subjects, and all of it is at least hoped to be novel - understanding all these things is impossible!
And there has been a trend on and off for the USPTO to grant patents pretty easily, reasoning that if somebody wants a weak patent, it’s not in the state’s interest to spend much time fighting them over it. The real challenge comes when somebody else infringes the patent and argues in court that it isn’t valid. All kinds of resources suddenly come to bear, attacking the patent. If somebody wants to write a weak patent and spend their money trying to defend it in court against a wealthy adversary, the USPTO doesn’t necessarily appropriate much taxpayer money toward the project of saving them from themselves; a simple warning that the application looks weak may be the appropriate expenditure.
(Not a lawyer, just a multiple patent holder here). There’s a difference between design patents and utility patents, but the advertisers mentioned in the OP may be glossing that over for the purpose of gaining an undeserved credibility. A design patent covers only the appearance of a product. It’s very much like a copyright or a trademark, and implies nothing at all about the product’s usability. Scam artists (I know they’re legal) who advertise their help in “getting your invention patented” take advantage of the existence of design patents and don’t necessarily explain the difference.
A utility patent, the “real” kind if you will, must describe an invention that is new, non-obvious, and useful. It can’t be useful if it doesn’t work.
You can get a design patent for a perpetual motion machine if you want to, and even advertise if for sale as “patented”. But you can’t get a utility patent for it, nor can you advertise it as actually providing perpetual motion.
This is a topic I actually gave a talk on at The Amazing Meeting 5 a few years ago. You can find the text here:
The short version is, yes, it’s supposed to work, but it’s very hard to challenge the utility of a patent application, and there are many loopholes and dodges that can be used to get around this requirement.
The patent statute requires that an “invention” be among other things “useful.” You are absolutely correct that even in subsequent litigation, this requirement is almost never made the successful basis of an invalidity challenge. Part of this is because the courts have taken a lazy position to the effect that “utility will be judged by the marketplace, and if someone’s infringing it, it must be useful.” This has been particularly unfortunate with the rise of so-called junk patents, where the patents don’t really mean anything, are being stretched to cover more than whatever useful invention they disclosed, etc.
As for the patent office rejecting based on utility – assume it will never happen and you’re largely safe. This is also the case for enablement and best mode requirements; the patent office simply can’t/won’t test them (which is one reason some question the statutory presumption of validity of an issued patent – it has demonstrably not been subjected to scrutiny under each and all of the requirements of the statute).
I had occasion to have lunch in the cafeteria at the U.S. patent office. The examiners are mostly young, don’t necessarily have advanced degrees or experience in industry. About a quarter of them appeared to be monoglot Vietnamese speakers. And . . . they’re government bureaucrats who are evaluated on throughput as much as/more than quality. Infer from that what you will as to the strength or even existence of a legitimate invention in any given patent, which should not be much.
That’s a nice explanation, but I think the real one is that the patent office is now supposed to fund itself through application fees, and thus needs to increase the productivity of examiners, which means they can’t be bothered to look for prior art. When I was in Bell Labs both my patents and all the patents of people in my group got initially rejected due to the examiner finding something claimed to be prior art. Once the paper was by another person in my group - our internal patent attorney loved that! My two more recent patents, no better, just breezed through with no response or challenge.
But the end result is all the same - there are a lot of junk patents out there, especially in software.
It’s worse than that – it’s supposed to fund itself and spin off money that the federal government uses for completely unrelated purposes. Google “patent fee diversion.” This creates all sorts of perverse incentives. Blame Al Gore in part (why not?) – as part of his “reinventing government” initiative, Dept. of Commerce emphasized “clearance rates” for pending applications (which led to more application churn in all likelihood). I read an article at one point that the patent office had a banner on the wall – “Our Mission: To Get Our Customers Patents.” Uh, no, not exactly.