I'm not sure of the legality but I definitely appreciate their product. This lawsuit seems odd because google themselves scrape content for their indexes. From what I see SerpApi is really just providing a machine interface that Google themselves refuses to provide users and visibility into SERPs which is also something that users should have available to them.
Google publishes how to control their bot - with robots.txt. They then obey those instructions. Google also takes some effort to not use all your bandwidth. Google isn't perfect, but they are at least making a "good faith" effort to be nice and this does count in court. Overall most will agree that in general what google does to allow people to find their website is worth the things that google is doing.
You can of course argue a lot of edge cases if you really want. For the most part I want to say "it isn't worth the argument". In some cases I will take your side if I really have to think about it, but in general the system google has been using mostly works and is mostly an acceptable compromise.
But their robots are enabled by default. So it is a form of unsolicited scraping. If I spam millions of email addresses without asking for permission but provide a link to opt-out form, am I the good guy?
At this point everyone knows about robots.txt, so if you didn't opt-out that is your own fault. Opting out of everyone at once is easy, and you get fine grained control if you want it.
Also most people would agree they are fine with being indexed in general. That is different from email spam where people don't want it.
Looking at SerpApi clients, looks like most companies would agree they are fine with scraping Google. That is different from having your website content stolen and summarized by AI on Google search, which people don't want.
The claim is SerApi is not honoring robots.txt, and they are getting far more data from google/more often than needed for an index operation. Or at least that is the best I can make out of the claim in court from the article - I have not read the actual complaint.
People are generally fine with indexing operations so long as you don't use too much bandwidth.
Using AI to summarize content is still and open question - I wouldn't be surprised if this develops to some form of "you can index but not summarize", but only time will tell.
Do you have an example of a court saying that violating robots.txt violates an existing law?
In Ziff Davis v. OpenAI [1], the District Court for the Southern District of New York found that violating robots.txt does not violate DMCA section 1201(a) (formally 17 U.S. Code § 1201(a), which prohibits circumvention of technological protection measures of copyrighted content [2]).
It's my understanding that robots.txt started as a socially-enforced rule and that it remains legally voluntary.
I'm probably just being naive though...