Copyright Tool Scans Web For Violations

The Wall Street Journal is reporting on a tech start-up that proposes to offer the ultimate in assurance for content owners. Attributor Corporation is going to offer clients the ability to scan the web for their own intellectual property. The article touches on previous use of techniques like DRM and in-house staff searches, and the limited usefulness of both. They specifically cite the pending legal actions against companies like YouTube, and wonder about what their attitude will be towards initiatives like this. From the article: "Attributor analyzes the content of clients, who could range from individuals to big media companies, using a technique known as 'digital fingerprinting,' which determines unique and identifying characteristics of content. It uses these digital fingerprints to search its index of the Web for the content. The company claims to be able to spot a customer's content based on the appearance of as little as a few sentences of text or a few seconds of audio or video. It will provide customers with alerts and a dashboard of identified uses of their content on the Web and the context in which it is used. The content owners can then try to negotiate revenue from whoever is using it or request that it be taken down. In some cases, they may decide the content is being used fairly or to acceptable promotional ends. Attributor plans to help automate the interaction between content owners and those using their content on the Web, though it declines to specify how."

Yeah

[Hijacked Public]

FTFA:

If it works, it's a fantastic invention

Its purpose aside, yes, it would be a fantastic thing to be able to scan the entire web and reliably identify the context and content of any specific media file type. Video, audio, image, etc. Particularly if it could identify purposely obfuscated content.

I'm in what is almost certainly a tiny minority of Slashdotters in that I actually create copyrightable material rather than only consume it. I'm again in the minority in that I think copyrights are a good thing and again in the minority in that I can separate out the purpose of copyrights and the evil actions of the legal arms of **AA companies.

Regardless, while scanning the internet for improperly used material sounds great on paper this will probably end up being as effective as finding water with a divining rod. The current tactic of locking down things at the hardware and OS levels will get more support from the media companies, not that they seem all that good at choosing tactics when the internet is involved.

search by hash?

[straponego]

Does Google allow searching by md5sum or equivalent? I'm sure they have the capability. While not as impressive as what this company claims, it'd also be more reliable for unaltered media files.

But it looks like the real "innovation" these guys are pushing toward is fully automated filing of lawsuits. I think that was in Accelerando, which is fantastic, and which you can download it free. [accelerando.org]

Re:Dupe

[AKAImBatman]

Pretty sure this is a dupe, or so closely related to an earlier story as to not matter.

It's not a dupe. (Unless you count anything that appears on Digg first to be a dupe.) However, it's also not the first story of its kind. About a gazillion companies have formed with the exact same business plan (save for the "hotness" at the time being digital music) and about a gazillion of those companies have failed to develop software that catches anything but the most obvious infractions.

Every so often, some RIAA/MPAA fair-haired boy manages to get funding for yet another attempt. He then fails miserably and the cycle repeats. You'd think the investors would learn. Unfortunately, they keep getting dazzled by the latest, buzzword-compliant technologies.

Re:Dupe

[Maximum Prophet]

Since copyright lasts a long time and doesn't depend on being defended like trademark, there will be some allowances "for promotional reasons" like this:

1. Leak copywritten material in easy to copy format to places where it will be copied
2. Watch viral marketing campaign take over
3. Profit
4. Wait 'til revenue falls
5. Find infringers using new scan tools
6. Sue them
7. Profit more!!!

Re:Yeah

[AdamKG]

and again in the minority in that I can separate out the purpose of copyrights and the evil actions of the legal arms of **AA companies.

Let's make one thing clear: the RIAA/MPAA lawsuits are not, in any way, shape, or form, an abuse, negative side of, misapplication or malicious use of Copyrights. They fulfill the role of Copyrights in the first place; they are the logical end result of a system that says citizens are allowed to distribute ideas (or expressions of ideas), then stop any further distribution of them.

The **AA lawsuits are ridiculous, yes. But the ridiculous part is not the litigation itself, it's the laws on which the lawsuits are brought under.

Re:Yeah

[kanweg]

I'm a patent attorney and no stranger to IP. Having said that, any IP law is, or at least should be, a balance to on the one hand freedom to operate (both for IP users and for IP creators) and on the other hand a means for compensation for IP creators. For patents, that balance is not there for patents on software. Also for patents, at least they last for 20 years max. For copyright, that balance is not there. And I'm curious to hear whether you think it is a good thing that whatever you create is still under copyright more than 40 years after you die.

Bert

Re:Raise.

[FooAtWFU]

You joke, of course, of course, but there are tools out there to detect when a bot is abusing your site and not following robots.txt. The usual technique is to hide a few links in your page, and also have these links blocked by robots.txt. When a user visits the link, they're banned from viewing the site. (Sometimes, a CAPTCHA-like utility for unblocking yourself is presented along with the 403 page, in the event that a particularly curious user manages to find the link and activate it manually.)

Re:i don't like robots.txt anyway.

[Anonymous Coward]

I did that a while ago to my universitys website and found an excel spreadsheet that was used as in inventory of chemicals, what room they were in, and the access codes to the rooms. No I didn't do anything with said information.

Re:search by hash?

[stivi]

Hm, what about computing checksum of the actual media contents? For example, compute checksum only for sound data in MP3 or image data in image files, ignore all other data/metadata. Usualy media files are containers for smaller objects or data streams... Resampled or modified contents would not be detected though.

What concerns me:

[botlrokit]

I'm bothered by this type of scenario:
 
"Dear [webmaster]:
 
It has come to our attention that your website, [sh*touttaluck.com], does not meet compliance in terms of a variety of copyright laws of the United States and other countries. Infractions indicated by our software include, but are not limited to:
 
Images created with an unregistered copy of Adobe Photoshop
Flash files created with an unregistered copy of Macromedia Studio MX 2004
PDFs created with an unregistered copy of Adobe Acrobat Professional
Content and structure created with an unregistered copy of Macromedia Studio MX 2004
Content and structure created with an unregistered copy of Microsoft Office Frontpage 2003
Images created with an unregistered copy of . . . "
 
...starting to see what I'm going with? I understand they're likely talking about copyrighted content such as prior art images or mp3 files, or maybe even damaging company secrets that are leaked by a whistleblower, and then redistributed for the intent of airing dirty laundry, but I'm thinking about the structure of a page itself. A person group or company who solicits a webpage to be created by a web design studio would now have to ensure that the studio itself is in compliance, or the products they use to create the pages are legal. That's where I get all nervous.