Google lobs lawsuit at search result scraping firm SerpApi
- Thread starter JournalBot
- Start date
This thieving cocksucking search/AI monolith is suing company a fraction of its size for doing exactly what it did to come into existence. Amazing.
Upvote
224
(256
/
-32)
ubercurmudgeon
Ars Praefectus
I guess I can see the business case. It is just that Google Search been so enshittified in recent years, that it is hard to perceive it as something worth "stealing".
Upvote
144
(147
/
-3)
Upvote
174
(175
/
-1)
One of the increasing number of legal battles you genuinely hope both sides lose.
Upvote
131
(132
/
-1)
Unlike Google, which would NEVER violate the choices of websites and rightsholders by, for example, reproducing website content in bullshit AI summaries no one wants...
Upvote
227
(231
/
-4)
The difference between "now and then" is likely that previously SerpApi was used primarily by academics and researchers, and now it is used by every wannabe AI startup. I've used it for proof of concepts and it does work well, but I've always assumed eventually Google would sue it out of existence.
I'd like to think there's a market out there for a third party like Kagi to offer a decent alternative paid search index. (They have closed beta API access.) Because the huge players either hoard their own or just strike a deal that nobody else can afford.
I'd like to think there's a market out there for a third party like Kagi to offer a decent alternative paid search index. (They have closed beta API access.) Because the huge players either hoard their own or just strike a deal that nobody else can afford.
Upvote
44
(45
/
-1)
Ah Google, being a shitty hypocrite I see. Isn't your entire business at this point stealing content from websites.
Upvote
95
(97
/
-2)
It rarely happens, but both all sides are wrong here.
Google is a content thief, SerpApi is also a content thief. The companies hiring SerpApi for their plagiarism engines are also as guilty as anyone else using hired goons to act for them.
There's a popular Norm MacDonald meme that applies here.
Google is a content thief, SerpApi is also a content thief. The companies hiring SerpApi for their plagiarism engines are also as guilty as anyone else using hired goons to act for them.
There's a popular Norm MacDonald meme that applies here.
Upvote
57
(61
/
-4)
So bad for so many years that I stopped using it altogether. Haven't touched it in years. The last straw was it ignoring -"XYZ" style commands, as if it knew better than I did what I actually wanted to search for—at that point, it was a shitty suggestion engine, not a search engine.
Upvote
105
(105
/
0)
Google's complaints about SerpApi illegally using Google's data are probably valid. One hopes that someone else sues Google for doing the same thing and quotes directly from Google's filings in this case.
Upvote
88
(90
/
-2)
fellow human
Ars Praefectus
Even if this wasn't obvious window dressing for their real intentions, I don't see how this would make a difference in court. I don't think you can sue someone for doing something to someone else, especially when that something (disobeying robots.txt) holds no legal weight.
It seems to me putting your content on the open web is making a choice about who can access it: everyone.
Upvote
44
(44
/
0)
Well exactly! Search gives you what you want. Suggestions gives the clients control over what you get (ads!).
Upvote
25
(25
/
0)
EXTRA! EXTRA!! Large Sinister Cauldron calls small pot BLACK!! You'll read about it here first folks, unless you've already seen the content scrapped search results!! Give a dime, feed a starving Newsie! EXTRA! EXTRA!!
Upvote
42
(42
/
0)
Isn't it more fun to read an Ars article plagiarized into AI slop and split into 10 slides?
Upvote
23
(23
/
0)
Google does provide a search API; this is what https://developers.google.com/custom-search/v1/overview provides. The difference is that SerpApi uses various remote browsers for the query, so you can explore how results change with geography. I am not familiar with if executing searches via the API from different locations makes a difference, so it could entirely come down to scraping results instead of paying for the API.
Upvote
9
(11
/
-2)
Nah, advertising is at least half; stealing content is most of the rest.
Upvote
9
(10
/
-1)
Upvote
25
(25
/
0)
Wow... Think I misplaced me can of tears somewhere. Poor big company, that have profited on the back of thousands of others, ripping their webpages and making sure no-one ever visits them anymore.
But they somehow don't like it the other way around.
But they somehow don't like it the other way around.
Upvote
26
(26
/
0)
From what I can tell, this API isn't for the whole web, just your own website (or websites).
https://developers.google.com/custom-search/docs/overview
Upvote
22
(22
/
0)
Google wanted to present themselves in a more positive light and removed negative words. So they removed the "no".
Upvote
5
(6
/
-1)
Upvote
6
(6
/
0)
It seems to me that this comes down to copyright. Google scrapes websites and uses the result to populate its search database. It is allowed to do that based on a fair use interpretation that says the use is minimal enough to not violate the rights of the original author. I don't think Google can turn around and claim copyright on data that it is using under a fair use exemption. Only the original copyright owner can make that claim. Even with the data that Google is using under license, they still don't own the data. The only entity that should have standing in court is the owner of the copyright. The only data that Google should be allowed to bring a claim about is data they published themselves, and even in that case their would be fair use exemptions that would apply to the company harvesting the search data. The only claim that I think Google might have in court is the unauthorized access to their systems.
Upvote
20
(21
/
-1)
Very disgusting on Google‘s part and I’m sure Meta and OpenAI will follow suit…..
Upvote
-5
(0
/
-5)
So, you’re saying that if I do research based on publicly available data, my derived work can’t be copyrighted?
Sites can opt out of Google’s usage.
Google apparently cannot opt out of SerpAPI’s usage.
Upvote
-5
(4
/
-9)
Only the new additions and I don't think a search engine has any new additions to claim since all it does is index existing content as is. Probably they have a claim over their search algorithms that rank the results. I definitely don't see 'they are violating our copyright by violating the copyrights of other people' going very far - is what I would say if we were in a normal time and not hypercapitalistic fantasyland.
Not in any reasonable way or definition of the term can anyone who wants to be anything to anyone on the internet opt out of their eco system since they control 89% of the market. This is apparently a situation the US government is okay with given recent results.
That's also ignoring the other things they do like their very liberal interpretation of how robots.txt should work or paying hardware manufacturers to ignore alternatives.
Upvote
8
(9
/
-1)
I remember back in 2022, when I'd just been laid off, I applied to an opening at SerpAPI (I was applying to most anything that I met the qualifications for).
They said they'd review my application as soon as they could, asked for a link to my GitHub profile, and I never heard from them again. Which is a shame, because I never got a chance to ask some of the questions I had, like 'how legal is this, like on a 1-10 scale?' and 'is there a plan for when the sites you're scraping try to sue you off the face of the earth regardless of the actual legalities?'.
They said they'd review my application as soon as they could, asked for a link to my GitHub profile, and I never heard from them again. Which is a shame, because I never got a chance to ask some of the questions I had, like 'how legal is this, like on a 1-10 scale?' and 'is there a plan for when the sites you're scraping try to sue you off the face of the earth regardless of the actual legalities?'.
Upvote
9
(9
/
0)
Not to mention, wasn't there something some number of years ago about Google basically stealing content via summaries that reduced traffic to the actual sites with the info? Or am I hallucinating?
Upvote
15
(15
/
0)
SerpApi does not have it’s own index. As far as I can tell, they call Google for each request, stripping out all content but the links, and returns it as JSON. It’s a glorified proxy that adds nothing, it just steals the value delivered by google and re-sells it.
One doesn’t have to like Google, but at least their core business of surfacing information and directing traffic has merit. Sites do have a real option to opt out, but every time publishers have tried, they always comes back promptly because their visitor numbers completely crashed.
One doesn’t have to like Google, but at least their core business of surfacing information and directing traffic has merit. Sites do have a real option to opt out, but every time publishers have tried, they always comes back promptly because their visitor numbers completely crashed.
Upvote
13
(15
/
-2)
Upvote
-3
(0
/
-3)