June 16, 2024

Response to Declan's DMCA Piece

Declan McCullagh misses the boat at least twice in his August 19th
column concerning the potential impact on computer science research of
the Digital Millennium Copyright Act [“Debunking DMCA myths,” c|net
News.Com, http://news.com.com/2010-12-950229.html].

First, the DMCA has two arms: one that prohibits devices that circumvent
copy protection, and one that prohibits acts of circumvention. The
research conducted by Professor Felten and his colleagues took place
prior to the time when the “acts of circumvention” provisions became
effective in October 2000. Thus, these provisions did not apply to that
research. However, there is little doubt in the legal community that
this research, and similar research, would be illegal under the “acts of
circumvention” provisions. Declan fails to recognize this arm of the
DMCA in his column.

Second, the chilling effect of the DMCA cannot be described by the
probability of conviction alone. One must also consider the magnitude
of the exposure if convicted. Because the “acts of circumvention”
provisions of the DMCA were not in effect at the time of the Felten
research, the probability of an adverse judgment was indeed small.
However, a group of highly respected legal consultants told Felten’s
employer that the cost of an adverse judgment could be truly enormous.
The combination of these two factors had a very substantial chilling
effect. (It is also the case that two individuals were likely to lose
their jobs if the paper was published. This illustrates the human
dimension of the chilling effect.)

Other issues, on which we shall not elaborate, include the
anti-dissemination provisions of the DMCA, and the civil (in addition to
criminal) provisions.

It is disruptive to the progress of research when scientists must first
consult with attorneys to determine if previously legitimate research
might be in violation of the DMCA. We are happy to agree with Declan
that “The DMCA is … an egregious law … and should be unceremoniously
tossed out by the courts.”

Edward W. Felten
Princeton University

Edward D. Lazowska
University of Washington; Co-chair, Computing Research Association
Government Affairs Committee

Barbara Simons
Co-chair, ACM US Public Policy Committee

Microsoft Does The Right Thing

Microsoft has decided not to try using the DMCA to censor a paper by former MIT student Andrew “Bunnie” Huang about security mechanisms in Microsoft’s Xbox videogame console. (See this announcment.) Good for Microsoft! If more companies behave like Microsoft did in this case, the DMCA would do less harm.

Let’s not forget, though, that the DMCA created a situation where Huang essentially had to ask Microsoft’s permission to publish the paper. He had to withhold his paper for a period of time and he had to spend sizable lawyer resources to get what the Constitution is supposed to guarantee him – his right to free speech. The DMCA is still a problem, but Microsoft kept it from being a bigger one.


I’ve been reading recently about the history of technology. That history is filled with lessons for policy-makers now. Here is one:

One of the reasons we should be wary about banning technologies is that it’s often very hard to tell what a new technology will be good for. It might take a decade or more for the true significance of a new technology to emerge, and even when we can guess at one application of a technology we’ll probably miss other important ones.

The telegraph is a case in point. When the telegraph was new, it was easy to see how the instantaneous communication it enabled would affect fields such as news gathering and finance. But the telegraph had a surprising number of applications.

Example: The telegraph enabled weather prediction. Previously, all you could do was watch the horizon for storms and wait for old-timers to sniff the wind and say, “Reckon it’ll rain.” With the telegraph, you could find out about current weather conditions in other places, build a weather map, look at the patterns of wind and pressure, and figure out what tomorrow would be like. This had a profound impact on people’s lifestyles, not to mention its benefits for agriculture and for safety.

Every new, general-purpose information technology has unexpected, valuable applications. If we ban such technologies or hamper their development, we’ll never know what we’re missing.


Lawyers use the term “self-help” to refer to all of the little steps people take to protect themselves. Locking your bicycle is self-help – even though it would not be necessary in a world where everyone obeyed the law, it’s a good idea in the real world. Fences and burglar alarms are self-help too.

Self-help is often a cheaper way of protecting our rights than relying on the legal system. Rather than having an army of cops to watch all of our bicycles, it’s cheaper for each of us to buy and use a lock. But self-help is only allowed in defense of your own rights – you can’t put a lock on my bike.

Some of the recent anti-tinkering laws claim to be nothing but self-help for copyright holders. That’s what some people say about the DMCA, and about the Berman-Coble bill. If those laws merely allowed copyright holders to enforce their legal rights, there would be no problem. But they go much farther.

Glenn Otis Brown, in an article in The New Republic Online, offers a helpful analogy:

Let’s say you extend the fence that runs around your yard across your sidewalk. If little Jonny hops that fence, you can’t call the cops on the pretext that he might steal your lawn furniture. The principle is simple: you can’t make an exclusive claim, through technology, on a public good (access allowed by the sidewalk). But the DMCA seems to say that the rules in the digital context are different. If little Johnny decrypts your DVD so he can watch it on a different kind of computer–something that ought to be protected under fair use–you can call the cops on the pretext that he might pirate your movie.

The analogy is an interesting one. If you fence off the real sidewalk, society will let Jonny off the hook for jumping the fence (as long as he stays on the sidewalk). But society will do more – it will condemn you for building the fence, and it might even send the sheriff to cut it down.

Some people like to argue that “intellectual property” is just exactly like real property and should have the same protection. Let’s all remember that the law protects the sidewalks too.

British Bill to Ban Mods to Cellphone ID Numbers

The British Parliament is now considering a bill that would make it illegal to change the IMEI number on a cell phone. Each phone has a unique IMEI which it uses to identify itself to the cell network; it’s like a serial number for the phone.

If you report your phone stolen, the cell operator blacklists your phone’s IMEI, effectively shutting off the stolen phone. But if the thief can reprogram the IMEI number, then he can keep using the stolen phone. Is this enough to justify banning the practice?

There are two questions we should ask ourselves in evaluating this kind of ban.

First: Is it necessary? Presumably there are already laws against stealing cell phones and against using a stolen phone. The bill only makes sense if those existing laws are somehow deficient. If the deficiency is that the existing laws carry insufficient penalties, then those penalties can be beefed up. If the deficiency is that the existing laws are hard to enforce, then we would still have to ask whether the newly proposed law is somehow easier to enforce.

Second: Does the practice that would be banned have significant legal uses? It is hard to object to a ban on practices that only criminals engage in, but if there are legitimate and legal uses of it then we should be much more reluctant to ban it. And we really have to try our hardest to think of legal uses – it is all too easy to ignore unusual or novel uses that would turn out to have great value in the future.

Based on my very limited knowledge, it looks like the ban on IMEI modifications may pass this two-part test, and so may be good policy. But I’m happy to hear any counterarguments.