Can P2P Filter Copyrighted Content?

scubacuda writes "DRMwatch reports that technologists acting on behalf of porn publisher Titan Media reported to Congress that P2P networks could (if they wanted to) use "fingerprinting" (aka "hashing") to detect copyrighted works and then filter them with the "spyware" installed on all nodes in the network."

Doubt it.

[BassZlat]

It is possible only according to the suits in the government. The p2p traffic accounts for ~2/3rds of the internet traffic nowadays, so unless you have an echelon-type system good luck!

(and that is not counting all the anonimity-protecting nets such as freenet [freenetproject.org], MUTE [sf.net], and the new i2p (don't remember link, sorry).

If it was truly peer-to-peer...

[vasqzr]

Couldn't it NOT be shut down?

Just like with Napster, there's a core that they can shutdown and be done with it. Are any of the popular P2P networks truly independent?

won't work

[Dreadlord]

filtering files based on hashing values won't work, especially for audio and movie files, you can always modify the file a bit, add a black frame to the beginning of the movie for example, so the hash value changes, and the file passes the filter.

Another ridiculous measure

[cavemanf16]

Wow, so now all the Divx rippers will have to chop a few frames off of each divx they rip so each hash is different. Companies should really stop worrying about what their customers do with the materials they have purchased and figure out a way to actually encourage them to purchase said materials in the first place. And no, I'm not just talking about pr0n, but CD's and DVD's in general. If it's a quality movie or CD I'll buy it because I know I'll want to watch it over and over and add to my 'collection.' I've spent more on Peter Jackson's works in the past two years than I have on any other media combined. (at least that I own... not counting all the Blockbuster rentals)

I mean seriously, how much money is Blockbuster making right now renting movies (some of which get ripped by the Divx kiddies 'cause they have way too much time on their hands) while the music industry bemoans their inability to sell records like they did in the late 90's?

Yay.

[elmegil]

Glad to hear Congress is listening to and believing sleazeballs from the porn industry blowing sunshine up their collective legislative butts. It's a shame we can't make congresscritters refer to an unbiased (hahahahahaha) technical agency who can tell them when these kinds of things are full of it.

I guess someone forgot to tell them

[Anonymous Coward]

P2P technology is worldwide, what is illegal in one place might be perfectly legal elsewhere, good luck trying to enforce it

of course the USA can have their own crippled P2P, the rest of us in the other 191 countries and 95% of the worlds population shall just carry on

you have to laugh at the stupidity of americans sometimes

Let's say you install the spyware...

[YinYang69]

If they use md5sum hashing, which the RIAA has already admitted to, all I have to do is change the comment entry in the ID3 tag of an Mp3 and I have a brand-new hash that they'll not be able to identify. That is unless they download it, test it for copyright (listening to it), and then add that hash to their md5sum DB.

But I can change my ID3 tags all day. Can they match me (hypothetically, of course ;)) md5sum to ID3? I highly doubt it.

This won't work

[Omnifarious]

There are systems by which the network cannot possibly detect whether material travelling over it is under copyright or not. Freenet is an example. Everything that goes over the network is encrypted. Nodes may not necessarily have decryption keys. There is then no way for a node to recognize a particular work.

time will tell

[bbowers]

So how long before the come up with a solution that actually works? If they do there will be ways around it of couse... our file sharing went down at school once for a few days and all you had to do was walk down the halls in the dorms and yell asking if anyone had such and such software/movies/music/porn, someone would stick their head out the door and you'd go burn it or run a cable down the hall

Will AD-Aware become a circumvention device?

[Ilex]

detect copyrighted works and then filter them with the "spyware" installed

So under the DMCA AD-Aware and all other spyware removal tools will be illegal as they could be used to circumvent DRM.

Sounds like a ploy by the pr0n industry to install more crapware on our pc's.

Come to think of it *nix will be illegal too as their spyware will only run under wind0ze.

Re:Didn't AudioGalaxy try this?

[R.Caley]

[...]they must know that works were not copyrighted.

This seems to open a possibility. Note I'm not saying this is a good idea, just that it seems like it might be a more workable system than most proposals:

Set up a public/private key infrastructure. If the content producers are losing as much as they claim they should be more than willing to pay. Anyone can have a key if they verify who they are to a reasonable level (eg by supplying a credit card number).

Now, we can have a rule that a client must only distribute a file signed with a recognised key.

I'd bet there are far fewer people willing to jump through these hoops and nail their credit history to their assertions of what is theirs or pubic domain than there are songs and pornographic images in the world, so this should be a more tractable problem than the finger printing.

Any client recieving a non-signed file reports the sender to the men with the big sticks with nails in.

Since there are relatively few people with the skill and interest to create hacked clients, and since such hacked clients should be reported if they are ever seen by a single legitimate client, it shouldn't be impossible for the MwBSwNI to keep the population of evildoers down. Especially if the punishment for distribution of such a client is suitably dramatic and well publicised.

Re:This *is* possible...

[0123456]

First, of course, your algorithm has to recognize that it's music. That leads to numerous obvious ways of avoiding such filters, in order of sophistication:

1) Rename the file from .mp3 to .txt, etc.
2) Put the file in an archive of some kind (.zip, etc)
3) Encrypt the file.

So the more sophisticated your scanner might be (e.g. checking file type is trivial, extracting files from an archive is easy, breaking encryption is hard), the more sophisticated the workaround becomes. Eventually the only way to break the filter-avoiding measures will be to have a human sitting there manually checking all the files they can find on the network.

There's an Easier Way

[StormyMonday]

Just have the pR0n suppliers encode a serial number in each copy of each video they sell. Then, if a copy got illegally distributed on the Net, they'd know who to go after.

A big job? Yes. But so is the "fingerprint database".

And this way, they'd be responsible for their own content, instead of requiring Big Daddy Government do it all for them.

Since they claim to be losing billions of dollars to "piracy", it should certainly be worth their while to charge a few bucks more for each video in order to increase their sales by (according to some numbers I've seen) an order of magnitude.

Re:It'll never work

[yoha]

This is wrong. There is something even more valuable than music and movies that is exchanged digitally every day, and that is money. If there were no way to protect it, then the banking system would have crumbled already. There are obvious differences between copyrighted material and money, but for practical purposes, they are both just bits of data exchanged between machines. If protection can be done for money, then it will be done for all other valuable data. Copyright protection will happen within 5 years. Current content which has been produced on CD's will be grandfathered into piracy, but new content will not.

Re:If it was truly peer-to-peer...

[johnos]

The argument makes sense, except there are legal and business considerations. The "common carrier" protection of traditional information movers like the phone companies will likely prevail eventually for data providers as well. The "safe harbour" clause of the DMCA was an attempt to head that off. However it probably won't work. There are a lot of big corporations that like being common carriers in fact if not in name. The phone companies, the backbone providers, Fedex. None want a serious precident that might make them liable for the content they move. They do want to keep the pipes independent and they have the money and lobbying power to slow down or stop any attempts to make them othewise.

Your local ISP may be intimidated by court orders or nastly letters from some lawyer. Verizon and SBC aren't. They want no part of the copyright wars because of the expense and potential for customer churn. The boys over in legal don't generate revenue and bandwidth is part of the cost of doing business. As long as one is willing to put up with the bandwidth from P2P, they will all have to put up with it.

I don't trust big corporations except to do whatever they can to look out for their own interests. As long as their interests and ours coincide we have a measure of safety.

Re:This *is* possible...

[kryptkpr]

I don't buy this.

Napster only ran the search servers. Files were still transferred peer2peer. So how could this technology "recognizes music by the acoustic properties of the audio itself regardless of how it was recorded, encoded" when the actual music is never seen by the servers, only the filenames? (Which was exactly how napster actually filtered.. by filename, the only information they actually had on the file, other then size).

And it's doubly nuts because it won't work.

[Ungrounded Lightning]

Of course it's doubly nuts because your proposal is trivial to beat. Add an extra random-sized bit of silence / blackscreen at the start and end. That changes the file size and shifts the hashed regions, causing all the hashes to come out different.

Trivial / hypothetical

[Anonymous Coward]

> Can P2P Filter Copyrighted Content?

Well, the Internet is based on TCP/IP, which is a P2P network architecture.

So if the Internet can filter content, then by definition, so can P2P. It is trivially obvious that the technology can support this.

> P2P networks could (if they wanted to) use "fingerprinting" ...

Well, people could (if they wanted to) respect the copyrights of rich media conglomerates.

You can ask all the hypothetical questions you want. But is there anyone who seriously believes that people want their P2P networks to be crippled with DRM?

This topic seems pointless to me. Is there any substance here for those of us who choose to focus on the feasible?