The problem with copy prevention: Why copy prevention won't work

By J.B. Nicholson-Owens (jbn@forestfield.org)

Tuesday, April 30, 2002

This essay aims to explain why copy prevention doesn't work and debunk some of the myths surrounding the reasons why electronic copyrighted items are encumbered with copy prevention at all.

Historically it hasn't worked

Personal computers first became affordable and popular in the 1980s. Some proprietary software developers tried to release copies of their programs so that only certain copies would function, namely the copies they sold. Software proprietors often believed that they would lose money if they let people share the software. The intention was to either stop people from making copies or to rig the software so any copies would not function properly.

Virtually all of their efforts failed. Some hackers (clever users seeking knowledge, not computer users doing unethical things) became adept at disabling or working around the locks that prevented them from making functional copies. Copy prevention schemes were easily worked around on early computers like the Commodore-64 home computer.

It was fairly easy to load the entire program into the computer's memory then write the content of the computer's memory to a file on a diskette. Then anytime you wanted to run the software you merely reloaded this file back into memory. Two effective copying products called "Isepic" and "Snapshot-64" did this for you; you would load the software normally from the original diskette, flick a switch on the Isepic or Snapshot-64 cartridge, and the cartridge would save the contents of the computer's memory to a file on another diskette. Copies became easy to make and easy to use.

Other successful copying efforts involved learning how the program tried to keep users from making copies. One copy prevention scheme relied on deliberately-placed electronic errors on the original software diskette. If you loaded the software from the original diskette, the software would check for these errors and only run if the disk errors were present. After a short time other software manufacturers became adept at making copies by reproducing these errors with ordinary home computer disk drives. This gave users another way to easily make copies of their software.

More imaginative anti-copying efforts involved the use of physical gadgets called "dongles" to prevent unauthorized duplication. Instead of looking for errors on the software's media, the software would look for the presence of the hardware dongle installed on your computer. If you installed the dongle before you loaded the software, the software worked. If you did not install the dongle, the software would not load or run. Each software package employed a different dongle. Users were expected to keep track of each dongle for each program. It's easy to imagine what a hassle this was. After a short time hackers learned to locate and delete the section of the program that checked for the dongle.

Today proprietors are trying software "keys". A software key is a special message (like a password) only handed out to legitimate owners of licenses (typically only the people who bought the software). The idea behind a software key is this: even if everyone copies and distributes the software, not everyone will have the key. So the software will only work for the select few that have both the software and the key. Hackers quickly learned to disable the section of the program that checks for the software key. Hackers also learned to make key generator programs (or "keygens"); programs that will create any number of valid keys.

The best response: meeting the challenge of the environment

Free software does away with all of this. Free software can be shared legally so there is no need to think about copy prevention rubbish. You simply obtain a copy of the free software program and run it whenever you want wherever you want. This is a far superior response than any kind of copy prevention because it grants users the freedom to build a community around the software. Software developers can build profitable features on top of that community—selling warranties and support are two popular ways to make money with free software.

Users want to do business with vendors that don't treat them like criminals.

The essential challenge of copy prevention is its Achilles heel

Ultimately all copy prevention boils down to one issue for the protectionistic: getting the computer to distinguish between allowable copies and disallowed copies. Allowable copies would hopefully include backup copies (but that's by no means certain). Disallowed copies would be any other copies, making a copy to share with a friend, for example (even though this kind of copying should be encouraged, not restricted).

For a computer, all copies are the same. A computer cannot determine the intent behind making any particular copy. Thus any copy prevention scheme is bound to fail. It's only a matter of time until it fails.

Digital locks fail for some of the same reasons physical locks fail: no lock is perfect. Every lock can be tricked into allowing entry with a forged key. Every lock can be bypassed.

If a key is somehow coded to a particular user, you discourage sharing the key and encourage bypassing the lock entirely. Some people believe that by "streaming" digital audio they can prevent other people from copying it. But it is easy to obtain a short audio cable to go between the audio output and input jacks. Plug in the cable, start the recording software and the playing software simultaneously and wait for the audio to be recorded. This is the high-tech equivalent of holding up a tape recorder to a speaker to record songs off the radio. There is also software to achieve the same effect without using the cable at all.

Ramifications of forcing copy prevention

There are startling ramifications of bills like the CBDTPA (neé SSSCA). You lose the ability to legally make backup copies. As the price of audio CDs goes up, you'll value having backups. Backup copies of anything are a valuable protection of your investment.

The CBDTPA reduces your right of fair use by restricting copying to what the publisher will allow. The publisher believes it is best to restrict all copying. But your ability to fully participate in a democratic and free society requires allowing you to make copies of small segments of copyrighted work without asking permission to make the copy. You need to be able to make unauthorized copies to be able to criticize, base new works on, and analyze copyrighted works in everyday life, whether your job encourages such activity or not.

So, how do we get computers to detect a user's intention for making a copy? We can't do it at all. Computers are dumb machines capable of performing a small set of mathematical operations very quickly. That is all computers have to offer. Computers are not capable of understanding law, politics, or ethics. Simulated understanding is unacceptable, we need genuine case-by-case analysis that only another human being can provide. We also need laws that protect our ability to share information and protect unpopular speech.

What's the harm in sharing?

Unauthorized duplication isn't necessarily a bad situation for copyright holders. As mentioned in a footnote in the landmark Diamond Rio decision, it's not clear copyright holders who sell copies of licenses to their copyrighted works are losing money when someone makes unauthorized copies to share with others.

Some arguments from copyright holders:

Unauthorized duplication results in lost profits.

Only if the copyright holder can prove the unauthorized duplicator would have paid the publisher for the copy they couldn't get any other way. Read that again slowly, there are three statements in there. All three of these conditions must be met to show a genuine loss:

1. The burden of proof is on the copyright holder.
2. The copyright holder has to prove the unauthorized duplicator would have paid for the copy if they couldn't get the copy any other way.
3. The copyright holder has to prove the unauthorized duplicator would have paid the publisher for this copy.

When copyright holders try to convince you they can answer all three points, don't forget: You can't lose money you never had.

Often those accused of widespread copyright infringement (colloquially known as "pirates", even though that is a propaganda word) have no intention of paying for the copyrighted item no matter what its price.

Making copies is a valuable service whether you distribute unauthorized duplicates or you spread knowledge of how to make an unauthorized duplicate. Backup copies are almost never provided by copyright holders (and such copies need to be backed-up themselves).

I had to raise prices to account for the loss to unauthorized duplication.

No, you chose to raise prices. This argument assumes the statement in the first argument is true, even though nobody has ever proven they actually lost income to unauthorized duplication. The main problem in raising prices to account for "piracy" is that the loss-of-income argument goes away. If you raise prices to account for what you believe you've lost, you're no longer losing that money.

After all, you're offsetting that "loss", so you aren't really losing much if anything.

Also, consider who is bearing the impact of this price increase: the consumers who have bought your product. That's right—by fighting this mythical problem of "piracy", you're driving away your customers; you're making it hard to justify buying another product from you because your prices are high. Perhaps these people will seek unauthorized copies in the future.

In other words, you're creating the problem you're trying to fight.

All the proceeds from sales of my copyrighted item should go to me.

Not at all. There are legitimate alternatives to buying a copy of a license and there are important social issues to consider. These alternatives make it clear why it is important to think in terms of obtaining legitimate licenses and not necessarily buying legitimate licenses.

First sale doctrine lets you sell your copy of a work. Assuming you obtained your copy legally, you can resell it. Lots of people make money reselling used goods. Therefore, first sale doctrine sets up a potentially long serial transfer of ownership where the copyright holder is likely to be paid only once.

Society values this situation because it helps people obtain access they wouldn't otherwise be able to afford. Libraries are celebrated for loaning copyrighted works to all patrons at no charge. Garage sales and second-hand shops make goods available inexpensively. In the end, a free society works best with an informed public. Reselling and loaning works are necessary to keep the public informed.

Think about what these less expensive copies mean for your future business: these copies help sow the seed that drives a consumer to become your customer.

I'm only defending my property rights.

It is a serious mistake to think of copyright as a property right. The heart of copyright is ideological not physical. Copyright in the US is best thought of as an exchange, not an item. Contrary to what people commonly call the union of copyright, trademark, and patent law—"intellectual property"—these are three completely separate areas of law (that can conflict). Copyright is about giving authors an incentive to publish more work, not granting infinite power to prevent others from copying works.

Finally, consider this excerpt from Siva Vaidhyanathan's excellent book "Copyrights and Copywrongs: The Rise of Intellectual Property and How It Threatens Creativity" (ISBN: 0-8147-8806-8):

The goal of the entire copyright system should be to recognize the pernicious repercussions of restricting information, yet to reward stylistic innovation. To envision the best possible copyright system—one that would encourage creativity and democracy—we must revise our notion of intellectual "theft." You cannot "steal" an idea, a style, a "look and feel." These things are the raw material of the next move in literature, art, politics, or music. And using someone's idea does not diminish its power. There is no natural scarcity of ideas and information. To enrich democratic speech and foster fertile creativity, we should avoid the rhetorical traps that spring up when we regard copyright as "property" instead of policy. We must also rediscover, reinvent, and strengthen the idea-expression dichotomy. And we will be able to have a more informed public discussion about the purpose and scope of copyright.