Fair Use and the DMCA Triennial Rulemaking

On Monday the United States Copyright Office released the results of its triennial rulemaking on section 1201(a)(1) of the copyright law, popularly known as DMCA after its 1998 enactment as part of the Digital Millennium Copyright Act.  The law criminalizes hacking of DRMs, or to use the legal term, TPMs (Technical Protection Measures).

The law requires that every three years the Copyright Office conduct a process for designating exemptions, which are classes of works protected by TPMs that can be hacked legally.   The Copyright Office accepts input from the public and then decides which exemptions to grant.  Granted exemptions only last until the next rulemaking.

The purpose of the triennial rulemaking is to help ensure that the DMCA stays relevant to new technologies and enables actual uses of copyrighted works that are fair and that the public demonstrates are significant.  At the same time, the exemptions that the Copyright Office can grant are quite limited in scope, and some ambiguity has already existed about that scope.

This time around, six of the 19 submitted classes of works were granted as exemptions.  The exemption that got the most publicity is, ironically, the one that shows best how TPMs and the DMCA have evolved beyond their original intended purposes: it is now legal (at least in copyright law) to “jail-break” Apple iPhones and iPhone apps.

Another exemption granted by the Copyright Office pertains to short clips of videos within protected DVDs.  The previous rulemaking made it legal to hack DVD encryption in order to get extracts from movies or TV shows and use them for educational purposes.  Now the exemption has been broadened to include using short clips for noncommercial purposes, period — as long as the use is for criticism or comment.

DMCA does not have a blanket exemption for “fair use of a copyrighted work,” or put perhaps more appropriately,  “works that, when protected by a TPM, cannot be used by the public in a manner consistent with Fair Use as defined in 17 USC 107, provided that such use is consistent with 17 USC 107.”  Copyright advocacy groups have tried to get such a broad exemption approved, and there were attempts to pass such an exemption into legislation, but to no avail.

Instead, advocates have had to focus on narrow subsets of fair-use issues to get exemptions.  Another granted exemption, for example, is software protected by dongles (physical devices that must be attached to your PC when you want to run the software) where the dongle doesn’t work anymore and is too obsolete to be fixed. This one carried over from the previous rulemaking.

What results from this process is, at least for protected content, a list of uses that are presumptively fair — albeit one that is patchy, incomplete, and ephemeral.  But in other words, DMCA does what the actual Fair Use law was not designed to allow.  The Fair Use law was designed to give courts flexibility to interpret cases brought before them according to guiding principles, not “lists of uses that are fair.”

Meanwhile, last month at the Copyright & Technology conference, I moderated a panel on Fair Use in the context of news publishing.  The panel included Sri Kasi, general counsel of the Associated Press, which has been introducing technology to detect uses of its vast output of news content online and make at least a first-cut determination of whether such uses are licensed or “fair.”

I had an “a-ha moment” on this panel which relates to my observation above about the Copyright Office’s DMCA rulemaking.  I had been asking the panelists about the possibility of making the Fair Use statute more fact- and rule-based, so that it might be possible to imagine a “Fair Use machine” that, say, a user-generated content website could use to decide whether to allow upload of content and under what conditions.  (For example: here is a piece of a news story.  Is the surrounding writing  criticism or parody?  Does it attribute the quote properly?  And so on.)  As everyone who studies the subject knows, such a “Fair Use machine” is inconceivable under current law.

The panelists (being lawyers, not engineers) didn’t think this was such a great idea.  There was some discussion among the audience about industry conventions, such as the 30-second rule about music sampling for promotional purposes.  Then came the a-ha moment: if there are widely-used industry conventions, then why do such rules need to be enshrined into law?  Perhaps the law really should be reserved for only those cases when the parties can’t agree on such rules without going to court.

But my point is a larger one, and it’s bolstered by the machinations of the Copyright Office’s triennial DMCA rulemaking.

Technology governs content to a greater and greater degree nowadays.  Its efficiency in letting ordinary people create, move, and copy content is breathtaking.  But when there’s a dispute over copyright, especially one that isn’t decidable by industry convention, the deciding factor is a legal system that is slow, expensive, and often capricious in its outcomes.  This is a problem that technology can help solve.

The exemptions granted in the DMCA rulemaking tacitly acknowledge the pervasive influence of digital technology.  And while they often fall short of being machine-readable rules (e.g., how short do “short portions of motion pictures” have to be to qualify for the exemption, and what exactly is “criticism or comment”?), they move in that direction.  That is, they give people guidelines on what’s fair and what’s not fair that are based on facts and circumstances of use, and that should not normally require hiring a lawyer or going to court.

An even better example is the exemption for hacking e-book DRM to enable the e-book reader’s read-aloud feature, so that the sight-impaired may enjoy the e-book.  That’s another exemption that the Copyright Office carried over from previous rulemakings.  That establishes “using read-aloud features s in e-book readers” as a quite straightforward fair use.

In any case, it seems to me that the Copyright Office triennial rulemakings represent a step — perhaps a small and tentative one in the grand scheme of things — towards avoiding what I have called a trap door into the legal system for fair use determinations, which is a chill on innovation in its own way.

The Copyright Office DMCA rulemaking process must, by definition, be revisited regularly — unlike legislation, which is notoriously time-consuming and unreliable.  But despite my a-ha moment at the conference last month, I still say there must be a way to eliminate the trap door, or at least make it smaller.  Perhaps expanding the scope of the DMCA rulemaking is the best place to start.


  1. When I read the changes, my first thought was this: Do the changes here do anything to discourage anyone at all (acting as a “noncommercial” videomaker) from renting a brand-new DVD release (lawfully acquired), breaking it into 10-minute segments (short portions), and posting the individual segments to YouTube with any subtitle commentary track (criticism and comment)?

    Or will Google/YouTube become the gatekeeper that decides what sort of artistic use or commentary merits playback on the internet?

    There’s a really well-done multi-segment YouTube commentary dissection of George Lucas’ “Attack of the Clones”, and I think that this change would really encourage more grassroots media projects similar to that (which is good).

    But the way I read it, the change might also open the gates for an awful lot of media trolls to pillage other peoples’ content for whatever thrills and shock value they can generate.

    Hopefully I’m wrong about that.

  2. Nicholas Sheppard · ·

    I’ve long wondered if the whole debate over circumvention devices is missing the point, since it seems to propose that fair use be available (or not) depending on the current state of an arms race between DRM makers and DRM crackers, rather than any meaningful socio-economic goal.

    Rather than say that blind people are permitted to crack DRM systems in order to use screen-readers, for example, why not just require that DRM systems permit the use of screen-readers in the first place? Is the goal of the policy to enable the use of screen-readers, or to encourage blind people to take up cryptanalysis?

    As Bill suggests, perhaps there’s a role for industry conventions here since each industry will have its own set of uses that might be considered “fair”. One can imagine a code of conduct for rights-managed content that set out the maximum number of devices in a “family domain”, the usability of screen-readers, etc. for a particular market. But making all those things dependent on the existence of circumvention devices seems perverse.

  3. Nicholas,

    Thanks for your interesting comment.

    Whether you believe that there ought to be a law requiring e-books to support read-aloud functions relates to your view on the role of laws – whether they should only be created to address known (not hypothetical) market failures or whether some societal good should trump market considerations. Now bear in mind that industry conventions have little to do with societal value; they mostly just follow market forces. E.g. music sampling is allowed because it serves an important promotional purpose for record companies, not because of some value-based notion that consumers ought to be able to taste before they can buy.

    As for the connection between industry conventions and the necessity of circumventing DRM, of course you’re right. The Copyright Office rulemaking only concerns circumvention. I was just generalizing that to the more general case of establishing fair use rules, whether or not circumvention is involved.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: