Files containing ONLY a 1 or a 0 are copyright infringing according to google drive

[darknessblade]

Summary

Files on google drive containing only the Symbol 1 or 0 are breaking copyright law according to google-drive

 

Quotes

Quote

 Users were left startled as Google Drive's automated detection systems flagged a nearly empty file for copyright infringement.

The file, according to one Drive user, contained nothing other than just the digit "1" within.

 

My thoughts

Did Google hit a new low?

How the bloody Hell could a single Symbol be Copyright infringing? 

ALSO: Isn't scanning files for "Copyright infringement" not a severe breach of Privacy? [at least in the EU. according to the GDPR]

Sources

https://twitter.com/emilyldolson/status/1485434187968614411/photo/1

https://www.bleepingcomputer.com/news/security/google-drive-flags-nearly-empty-files-for-copyright-infringement/

[emosun]

Its clearly a glitch

[Mark Kaine]

2 minutes ago, darknessblade said:

at least in the EU. according to the GDPR]

if eu would have some balls they would just ban google altogether, it is clear alphabet inc does not care about rules and is only interested in exploiting peoples data and acting above the law.

Ive seen some "notifications" from google that actually made it sound  like they're the police, "you are required (by law*) to send us copy of your ID card or else…" basically…

 

*even though thats not at all what the law "requires"

[Franck]

8 minutes ago, darknessblade said:

ALSO: Isn't scanning files for "Copyright infringement" not a severe breach of Privacy? [at least in the EU. according to the GDPR]

GDPR is protection of data it has nothing to do with verifying if the content is legal based on the view law of the country you are in. Scanning for treat is actually covered and encourage as the data of the user is as secured as the weakest point. As an example making sure the user does not upload a virus which might affect his or other data is critical problem. Anyhow EVERY file send to a server is read to memory at some point. This is how the file get from point a to point b.

[Blademaster91]

16 minutes ago, Mark Kaine said:

Ive seen some "notifications" from google that actually made it sound  like they're the police, "you are required (by law*) to send us copy of your ID card or else…" basically…

That should be illegal if it isn't, i've never gotten any notification requiring an ID, but I wouldn't trust any of the large tech companies with my ID card.

 

[grg994]

FYI the amount of information in a file necessary to claimed copyright infringement is surprisingly low.

Here is a precedent with only 128 bits: 0x09F911029D74E35BD84156C5635688C0

https://en.wikipedia.org/wiki/Free_Speech_Flag

[Mark Kaine]

5 minutes ago, Franck said:

GDPR is protection of data it has nothing to do with verifying if the content is legal based on the view law of the country you are in

this seems true, but the law is really ambiguous here, they're not allowed to do this for "individuals" so how does this work? Everything anonymized, but if they find something actually illegal wouldn't they be required to inform authorities about this "anonymized individual"?

 

[poochyena]

16 minutes ago, Blademaster91 said:

That should be illegal if it isn't, i've never gotten any notification requiring an ID, but I wouldn't trust any of the large tech companies with my ID card.

 

But thats how ID cards work. if you don't trust them with your ID, then you simply don't trust ID cards.

[Mark Kaine]

13 minutes ago, Blademaster91 said:

That should be illegal if it isn't, i've never gotten any notification requiring an ID, but I wouldn't trust any of the large tech companies with my ID card.

 

The thing is, it should , but it isnt…

 

they are actually required to confirm age for age restricted content  (wheras it is highly questionable what age restricted content even is as that seems to be at the sole descretion of google, or the uploader for example) but there is no law that requires this verification through ID card, there are less intrusive options, which google promptly chose to ignore. And I agree, yes, *that* should  be illegal, they should be required to inform the user about all available verification methods, not just the one they profit the most out of.

[Franck]

16 minutes ago, Mark Kaine said:

this seems true, but the law is really ambiguous here, they're not allowed to do this for "individuals" so how does this work? Everything anonymized, but if they find something actually illegal wouldn't they be required to inform authorities about this "anonymized individual"?

 

It is not anonymized. It is protected / encrypted. A service might choose to anonymized but they have to have their headquarter somewhere it allow them to do so with the information so when the authorities try to get the information about the illegal content they can say "we don't have it as we don't need to comply".

 

If you put illegal data on a service and as GDPR you can request to remove everything they have about you (with some minor exception). If they remove information about illegal activity they are responsible for destroying evidence under some country law. I is very complicated and really case per case scenario as where the company is located, where does it service, where the user is located. Back when i was dealing with this kind of issues daily in the end it was reviewed by law department and they dealt with it.

[Forbidden Wafer]

Hmm, they banned a perfectly good forum in Google Groups... From a project that has been part of Google Summer of Code for years...
They took a few days to restore it.

At the same time, my poor email is piling up with spam. Their spam filters are complete trash these days.

[RejZoR]

If it was compressed file it's considered as decompression bomb because you can fit 50GB into a tiny few megabytes sizes file that will stuff all the subsystems if they try to analyze it.

[Franck]

16 minutes ago, RejZoR said:

If it was compressed file it's considered as decompression bomb because you can fit 50GB into a tiny few megabytes sizes file that will stuff all the subsystems if they try to analyze it.

you can't compress 1 byte tough

[wanderingfool2]

1 hour ago, darknessblade said:

ALSO: Isn't scanning files for "Copyright infringement" not a severe breach of Privacy? [at least in the EU. according to the GDPR]

Not really.  You are agreeing to Googles terms of service, as well from my understanding they are still letting you download and upload the thing flagged as copyright...just limited in sharing.  In this case, I suspect that either it's just a glitch in the algo (since it's 1 character it probably messes with some types of algos).  The only bad thing I could see about this is that they don't have an option to disbute the claim.  Then again, maybe someone has actually declared their ownership of the copyright of "1" and "0" and instead of it being an algo glitch it is someone who literally is laying claim to something non-copyrightable (which then they flag).

 

I suspect the reason they are doing this is to prevent what happened with the YouTube case.  ie. being sued by a copyright troll company and spending millions on millions to defend against it...or worse, when the courts determine that Google didn't do enough to try blocking it (look up GEMA v YouTube) [They didn't directly commit copyright infringement but by not acting against the users the courts ruled YouTube didn't do enough].

 

I mean ultimately, I think more and more services will start doing this kind of thing (or rather any that could be shared).  The reason being that copyright lawsuits are expensive, and the larger the company you are the more people will target you with lawsuits.  The way they currently have things setup, it seems as though it might be the start of a way to detect these issues without having to outright ban users (to protect themselves).

[GhostRoadieBL]

Unfortunately content search and content queries are part of the user agreement.

The algorithm will pick up false positives for partial code and full blocks of code so copyrighted things can't be shared.

 

This doesn't mean that there's a breach of privacy since even with the flag, it's a line of code in a file on a device Google owns. This is a major flaw in content ownership laws.

 

Do you own the file or does the possession (hdd) holder own the file?

 

If it's something you have creative licensing on, does the holder have to be informed about the the contents to be able to follow the ownership laws?

[Doobeedoo]

You can't use a number, it's mine.

[darknessblade]

32 minutes ago, Doobeedoo said:

You can't use a number, it's mine.

Me: Selects said number. CTRL-C

Me: CTRL-V the number into a text document

[emosun]

3 hours ago, Franck said:

you can't compress 1 byte tough

one bit*.  you can compress a byte

[jagdtigger]

TLDR
Encrypt your files before uploading them....

[Mark Kaine]

5 hours ago, Franck said:

It is not anonymized

but then its against EU laws.

 

i mean i didnt really mean to ask how they do it, more like how they get away with it.  And also pointing out the contradiction in these laws themselves. So either these services are allowed to spy on individuals data or they arent, kinda cant be both.

 

and i totally see why google in the end would say eff it and just do what seems safer to them even if it means a bit of breaking GDPR laws… 

[RejZoR]

12 hours ago, Franck said:

you can't compress 1 byte tough

Oh, it's single character. After reading it I got the impression it contained only "1" through entire it. Read it again and it says single number "1". Many 1's you can compress hugely because you just make 1 index and how many repeats in the index. Which means all of it. It's why that can expand into huge files after extraction. But yeah, not this one.

[Sauron]

14 hours ago, darknessblade said:

How the bloody Hell could a single Symbol be Copyright infringing? 

You could apply this reasoning to all copyright on digital media. It's all just numbers.

14 hours ago, darknessblade said:

Isn't scanning files for "Copyright infringement" not a severe breach of Privacy? [at least in the EU. according to the GDPR]

Depends on how it's done, besides your files on google drive may not contain any identifying information about you and even if they do it's not an issue so long as that data is not stored elsewhere without your consent.

[LAwLz]

14 hours ago, darknessblade said:

Did Google hit a new low?

No, it's just a glitch. It will be fixed.

 

14 hours ago, darknessblade said:

How the bloody Hell could a single Symbol be Copyright infringing? 

I mean, it could, but in this case it isn't.

 

14 hours ago, darknessblade said:

ALSO: Isn't scanning files for "Copyright infringement" not a severe breach of Privacy? [at least in the EU. according to the GDPR]

No, it's not. Nothing in the GDPR prevents your files from being scanned in this way.

 

8 hours ago, Mark Kaine said:

but then its against EU laws.

No it isn't. GDPR doesn't require all data to be anonymized.

 

 

I think a lot of people on this forum don't understand GDPR.

[Kroon]

 

So a bug?

[Franck]

4 hours ago, LAwLz said:

I think a lot of people on this forum don't understand GDPR.

Honestly i don't think any one does. The law was made to be confusing to be able fine as much people as they want. As a 25+ years professional developer i can tell that when you tell the client about complying they all go through their law department and every lawyer sees the law differently. We don't apply the law it the same way as it's per company basis. They all tell you you are wrong and they are right because their lawyers are right and you just a dev company that does this for a living. Wells anyhow this is not just GDPR that is like this. Some law firm don't consider your IP at home, name, last name, address as a personal information as they say it is publicly available informations.