Nude Photos Get Google in Trouble Over Copyright Law

[rustybrick]

I just posted about Google Image Search Loses Perfect 10 Photo Site Copyright Case at the SEW blog.

Perfect 10, an adult photo site, has proven to the court that Google's image search thumbnail copies are a violation of U.S. copyright law. This past Friday, U.S. District Judge A. Howard Matz, ruled that Google has violated the law "by creating and displaying thumbnail copies of its photographs."

Danny then noted to me, why didn't they bring up the robots.txt thing.

[Joseph Morin]

Hmmm...how much free traffic was Google sending them?

Reminds me of the absurdly ridiculous case of Ticketmaster vs. Microsoft where Ticketmaster didnt want Microsoft 'deep linking' to their concert venues - which would have made each venue a lot more money from the link popularity alone - although at the time link pop within an algo was just getting started but still, that lawsuit turned out to bite them in the butt because imagine where they would be now.

[michaelzimmer]

They didn't lose the case - this is just a preliminary injunction.

[rustybrick]

Blog entry titles...

[dannysullivan]

Yeah, got that title fixed now.

[Chris Boggs]

nice barry...trying to get this thread ranked for "nude photos" are we?

[mcanerin]

I find it a very interesting case, with potentially wide ranging results. Keep in mind that all that was decided here was that it was "likely" P10 would win at least some of their arguments. This was not a full blown court case, and the standards are much lower. It's not over yet.

Here is the crux of the decision. There are two major ways to display someone else's content on your site: 1) copying it from their server and displaying it from your server, and 2) displaying it directly from their server using frames or some from of an include or scraper.

Google argues that the test should be server based. What server is the information coming from? In the case of framed content, the content is coming from the original server.

P10 argues that the test should be the incorporation test, whereby the test is "what website does this information appear to be from to the user?"

The court acknowledges that both tests can be abused, but prefers the server test, since it's easier for webmasters to understand and preserves the interconnectedness of the web.

Frankly, I disagree with this. I think the court is focused on frames, and does not realize the implications of applying this to includes and scrapers. Further, frames do not always need to be visible. I believe the court is focusing too much on the particular facts of this case and is ignoring the wider implications. They keep talking about the grey bars separating the frames and the displayed URL, for example.

Ironically, the court sides with Google as to it's preferred test, then proceeds to use that test against them, in that they admit that the thumbnails are created and stored by Google, and therefore are a copyright violation under their own test. The major reason is that the thumbnails are so large ( up to 8 times a human thumbnail) that they are useful by themselves for the purposes of PDA's and cell phone wallpapers, for example. I would further add that many images on some sites are actually smaller than Googles thumbnail, and thus, there is no "thumbnail" at all - just image theft via direct copy. I fully agree with the court regarding this.

I disagree with the court in another area, though. They find that because Google makes a lot of money from Adwords/Adsense, that the image search is therefore commercial. Since image search does not display Adwords, they attempt to connect the dots by pointing out that the more useful and popular Google becomes, the more people will use it, and therefore that increased traffic will make them money.

I'm not sure I buy that. I don't think the halo effect on traffic of image search is enough to declare that the image search is commercial. By that definition, everything a company does that people like, including charitable donations, are "commercial" because they tend to build loyalty. I think that's going too far. Now if Adwords was displayed on the image search, that would be a different issue. I personally don't see a significant connection between image search and Adwords.

The only way it could be connected would be to show that people were making images available with the intent of driving traffic to their sites, which in turn sold Adwords. That's closer, but then the one doing all this is the site owner, not Google. There is no system that I am aware of that knowingly connects image search to sites with Adwords.

Just some (initial) thoughts,

Ian

[mcanerin]

<added>

Regarding Adwords, it appears my initial assumption was wrong - the court said there was no significant connection between Adsense and Image search.

I'm glad to hear that - I didn't buy it either.

Finally, the court basically held that Google was likely to lose due to the thumbnail issue, but would probably win it's argument that they were not responsible for *other* peoples copyright infringments that image search dug up.

Ian

Why would they only be targeting Google? What about http://search.msn.com/ you can also search and download Perfect 10 images. Must be old Bill pulling some strings.

[mcanerin]

I doubt it - it's a bad strategy to fight more than one battle at the same time (not to mention very expensive). It's best to choose a representative company with deep enough pockets, and go after them in a tightly focussed manner.

Much better to win a case, then take that winning case in hand and go after everyone else. Since the ruling is fresh and on topic, it's far easier, and most just give up rather than fight it.

Ian

[Mikkel deMib Svendsen]

I'm not sure I buy that. I don't think the halo effect on traffic of image search is enough to declare that the image search is commercial.

I am not sure I agree with you. Most (if not all) a commercial company do is done to make the business better - including contributions to charity (that is ultimately done to make people think they are "nce" and therfore like them them more, use them more etc).

I really don't see the logic that "stealing" is OK, if you only display and sell the stolen goods in an area of your shop where you do not sell other of your products (even though you have signs pointing to that department).

The search engines are highly commercial companies. If they do not think image search has any value to them they should drop it. If they think it has value they need to work out a deal with the copyright holders. You can't argue that it's OK to steal only because you claim what you steal have no value to you. If it has no value - then why steal it? ... I took your car, but it dosn't really drive that well, so I guess it's ok

[dannysullivan]

I need to read the ruling in more detail, but I don't recall that framing was an issue. I think the court was actually cool with that.

Google seems to have added a send image to phone feature recently. Perfect 10 sells downscaled versions of its photos that you can have on your phone. So Google was found to be directly having an impact on that for fee service. I suspect the injunction might be to prevent that particular send to phone feature from working for these particular photos.

I still remain perplexed about why robots.txt doesn't appear to have been an issue. I've got a question out to Google on that, but haven't heard back. If Perfect 10 doesn't want to be in the index, robots.txt would keep them out, problem solved. But in the Kelly case, he didn't feel he should be required to use robots.txt. Maybe that's the same here.

[Mikkel deMib Svendsen]

I still remain perplexed about why robots.txt doesn't appear to have been an issue.

I really don't think robots.txt should be an issue. If a act is illegal I don't think it should be OK if you did not tell people (or Google) NOT to do it.

Imagine that a militant group of feminists (just to use a crazy example) decides that unless I put a candle light in my window they will come and beat me up at night. Would that be OK too? After all, they HAVE issued a way for me to opt-out - I can just put up the candle light, right? How is this any different the robots.txt? What if I just don't want to have to deal with the candle light or the robots.txt but still want the law to apply (not to get beaten up - or to get my copyrighted material copied)?

In a case of copyright violation I don't think it is a valid argument that there is a way to opt-out of the violation taking place. The question should be if there is a violation or not in the first place.

[Chris Boggs]

seems to me that the simple solution would be for P10 to block the content from being indexed (hmm sounds familiar).

If a website places thumbnails out in the open, unless they protect the image, it can be donloaded to anyone's computer, and then onto a cell phone, n'est-ce-pas? If the adult content websites continue to offer free "preview" images and "sample videos" (I've heard about these too recently ), then I see no reason why G cannot index the content just like any other bit of information in its databases. (but OK G should adhere to robots.txt exclusions)

[Mikkel deMib Svendsen]

then I see no reason why G cannot index the content just like any other bit of information in its databases.

First of all, I think there is a very big difference between the way image search and web search works. Most people that use the websearch click the listings they want to see but I suspect that many users of the image search simple use the images as they are republished by Google - directly on Google.com without ever visiting the site.

Secondly I don't neccesarily think web search - the way it works today with caching and all, is neccesarily legal. Only time and real cases in all the juridictions the engines operate in will show. Caching and image search could turn out to be illegal in Germany - but legal in the US.

[vayapues]

First of all, I think there is a very big difference between the way image search and web search works. Most people that use the web search click the listings they want to see but I suspect that many users of the image search simple use the images as they are republished by Google - directly on Google.com without ever visiting the site.

Secondly I don't necessarily think web search - the way it works today with caching and all, is necessarily legal. Only time and real cases in all the jurisdictions the engines operate in will show. Caching and image search could turn out to be illegal in Germany - but legal in the US.

I often only look at the results on Google, and never actually go to any sites. As an example, yesterday, I could not remember for the life of me when the Triassic Period ended, and the Jurassic began. I searched "Triassic period end" and in the third result down, you see my answer, so there is not need to go to anyone's website.

I search this way, at least three or four times a week.

Similarly, I often use image search to find a product or other topic I am researching, when I am not entirely sure what the name is, but I know what the thing looks like. This saves me a lot of time. When I see what I want, I then go to the site, and read more. I don't have to click on dozens of sites with items that have a similar name, but are not what I need.

The vast majority of users are probably not savvy enough to swipe another person's image via Google Images. Of course all of us know it is not hard to take it, but most people are not very tech-savvy.

Now, if Google is offering a feature to send it to you, this opens up a whole new can of worms, and in this case, I would agree that they are violating copyright. (probably)

[vayapues]

Actually, the more I think about it, the more I am starting to believe that they would be more in violation (from a layman's point of view) regarding caching text, than images. (I will leave the actual legal prognostication stuff to those who know, Ian).

• They Monetize off of text search, where as they do not off of image search
• When displaying text, there is no reason to go to the users site, if you can already read what you are looking for either in the description of the site, or on the cache page.
• If you search via images, you absolutely have to go to the website, if you want to do any type of research.

As a publisher who makes much of my living on ad revenues generated off of articles written for my children themed websites, I should be far more concerned about text caching than about image caching. But actually, I am not really concerned about either.

By the way, Google Images is my 8th highest referrer, so certainly enough traffic to make it worth it to me.

[dannysullivan]

I really don't think robots.txt should be an issue. If a act is illegal I don't think it should be OK if you did not tell people (or Google) NOT to do it.

Until now, the last court case found image search was not illegal if it involved thumbnails and if there were disclaimers or something like that, at least in the US.

I think robots.txt remains very much an issue because of the other case, the caching case, where it was viewed as something a site owner could do if they wanted to opt out and effectively enforce copyright.

Of course, Google could always come back and argue that it's caching the photos

Anyway, there's still so much of this actual law about search to be tested. I remain surprised that robots.txt doesn't appear to have been raised as a defense. It would have been natural to do that. Not saying it would have worked, but it's the first thing you'd raise.

[mcanerin]

I think an issue is that calling something a thumbnail does not necessarily mean it *is* a thumbnail.

The general rule of thumb for fair use is that you use the minimum necessary to get the point across, rather than focusing on the maximum amount you can use.

At the point where the thumbnail is large enough to be a useful or desired image itself, fair use, IMO, no longer exists. At that point it's a copyright violation.

For robots.txt - I agree that it should certainly come up, but what if you are OK with Google indexing your images in order to facilitate search traffic, but are NOT OK with Google indexing your images to *replace* your search traffic?

Ian

[Mikkel deMib Svendsen]

In the Newsbooster case from Denmark robots.txt was in fact used but refused by the judge as not being relevant for the reasons I mentioned above. In fact, one of the winning newspapers in the case actually HAD a robots.txt file in place excluding all bots and Newsbooster respected it - but the newspaper still won! The last part I personally found a bit over the top. However, it's the only case on search here so far and as such I am pretty sure it would be used in other search/copyright cases here if they ever came up.