tl;dr Summary

The DMCA’s procedure for taking down allegedly infringing material is (ahem) nearly universally disliked. Rights holders can’t figure out why it’s so hard to take infringing material down, and consumers and service providers can’t figure out why it’s so easy to take obviously non-infringing down. Lenz v. Universal thus appeared a perfect test case for those who wished to clip the wings of the DMCA’s takedown procedure. When Universal saw Ms. Lenz’s video of her toddler dancing to “Let’s Go Crazy” in the background, it got the video removed. Although Ms. Lenz was able to get it put back up, she was angry enough to sue Universal under §512(f), which forbids knowingly representing that a work subject of a takedown notice is infringing. Her theory was that her video was so obviously a fair use, Universal’s takedown notice constituted a knowing misrepresentation. She knew, however, that she had a tough road, because at most all Universal had to prove was that it had a subjective belief that her video wasn’t fair use, a very low standard indeed.
At trial, she successfully fended off a motion to dismiss and a motion for summary judgment, in which Universal argued that it need not have considered fair use at all. After this point, this became all anyone knew or cared about the case: the abstract question of whether fair use must be considered before sending a takedown request. By summary judgment, she was able to prove that Universal didn’t explicitly consider fair use (though it considered some fair-use-like factors). The trial court held Universal was obligated to have at least considered fair use, and because Ms. Lenz had provided some evidence that Universal hadn’t, it would be up to a jury to determine whether Universal’s failure constituted a knowing misrepresentation.
Last week, the Ninth Circuit affirmed the trial court. Because of a change in Universal’s strategy, the decision was much focused on the nature of fair use, and whether it’s an affirmative defense or a right, or perhaps a flavor of affirmative defense. As I show below, this is a red herring. What matters is what “authorized by law” means in the phrase “not authorized by the owner, by its agent or by the law.” Ms. Lenz had been arguing that “authorized by the law” includes fair use, since a defense to something is a kind of authorization (in a weak sense). But Universal actually had a good comeback to that argument: there are actually several times where the Copyright Act literally authorizes use of the material and forces the owner to license the work. These are known as statutory (or compulsory) licenses.
In the end, it might not have mattered what Universal argued. It didn’t really appear the Ninth Circuit was in a mood to accept any of Universal’s arguments. And so, it appears that rights holders will have to consider fair use. But only a tiny bit. The subjective standard still applies, and with a little bit of education, rights holders should have little trouble spotting “easy” fair use (yes, there is such a thing100And I’m a bit of a hypocrite, because I myself like to play up how hard fair use can be.), and their good faith calls on other forms of fair use will not subject them to liability. For all the sturm und drang, the balance between rights holders and consumers has barely budged.
This analysis is based on the following documents:
* Universal’s Motion for Summary Judgment,
* Universal’s Appellate Brief,
* The Ninth Circuit’s Opinion.
* 17 U.S.C. § 512. We’re most interested in subsections (c) and (f).
By the way, the DMCA safe-harbor provision are something of a hobby of mine.

This seismometer measures changes to copyright law, starting in 1909. The big disturbance is either the 1976 Act or possibly the Sony Betamax case. Note that it's been a while since the needle has moved significantly.

This seismometer measures changes to copyright law, starting in 1909. The big disturbance is either the 1976 Act or possibly the Sony Betamax case. The smaller disturbance near the bottom is the DMCA in 1998. Note that it’s been a while since the needle has moved significantly.

DMCA Safe Harbor: Nobody Loves This Baby, Even When it’s Dancing

Some laws are unloved, but the safe-harbor provisions of the DMCA are actively hated by almost everyone. One might wonder why it hasn’t been repealed. That’s because the only thing worse than the DMCA safe harbors is no safe harbors at all. One might wonder, then, why it hasn’t been fixed. That’s because, though the DMCA safe harbors may be universally loathed, they’re loathed by different people for different reasons,101In terms of loathsomeness, the DMCA safe harbors are thus of a different quality from the Computer Fraud and Abuse Act, which is loathed by everyone except federal prosecutors, who actually want it expanded. so no one can agree on how to fix them.
The DMCA was passed in 1998.102And took effect in 1999. Hence the “M” for “Millennium.” Before then, if a copyright owner found its works being infringed online, its easiest path to success was to sue the company that hosted the site or otherwise provided the service that allowed the infringement. The service provider wasn’t liable for copyright infringement directly, but it was liable for contributory copyright infringement. In a perfect world, the service providers would eventually wise up and start policing their site for copyright infringement.

DMCA Safe-Harbor’s Rough Birth & Awkward Adolescence

There were four problems with this state of affairs, both of which had to do with the extreme asymmetry of legal resources. First, not all copyright owners could afford to even threaten a lawsuit. Second, back in 1998, most service providers didn’t have the wherewithal to police their sites. Innovative and interesting sites would be cut down before they could blossom. Third, even in 1998, the tendency was to view the internet as a Wild West or a bubbling cauldron of laissez faire economics and exchange of viewpoints. Fourth, the combination of digitization and the internet really was an unprecedented threat to copyright.
Congress decided to thread the needle between the interests of the service providers and the copyright owners. It created a process that made it much easier for copyright owners to remove unauthorized material but much harder to recover damages from the service provider—or, really, from anyone else. You’ve certainly heard of it: it’s the notice-and-takedown procedure.
The procedure is supposed to work like this: (1) The copyright owner finds something on the internet that it believes infringes its copyright. (2) It clicks on some identifiable link, such as “DMCA” or “Abuse,” that gives it the name, mailing address, telephone number, email address and even fax number1031998, remember? of a person who the service provider has specially designated to handle copyright-infringement complaints.104In theory, the copyright owner could look up the agent on the U.S. Copyright Office’s website, since the agent must be registered there. No one does this for a variety of very good reasons. (3) It writes a fairly simple “notification”105Not “notice.” No. It’s “notification.” to that person. (4) That person ensures “expeditiously” that the offending content is taken down, but first gives the uploading user a copy of the notification. (5) The offending material is taken down. (6) If the uploading user thinks she had a right to upload the material, she can send a “counter-notification” to the designated agent but must reveal her identity in the process. (7) Upon receipt of the “counter-notification,” the designated agent forwards it to the copyright owner. (8) If the copyright owner disputes the uploading user’s right to use the material, it sues the uploading user.
As you can see, the idea is the get the service provider out of the way. Either the material stays down, or the uploader pops her head up, and the fight is then directly between the copyright owner and the uploader.

The Takedown Procedure Might Not Be Quite What You’ve Heard

This is how it’s “supposed” to go. But it doesn’t have to. First, the service provider doesn’t have to comply with the notification. It could, if it chose, review it carefully and reject it. There is almost no advantage to doing so, mind, but it could. There is also a clear disadvantage to doing so: it risk liability for “contributory copyright infringement” for hosting the allegedly infringing content. So, naturally, 99.99%106Honestly, I’m just guessing with this number. It might really be 99.98%, for example. of all notifications result in a takedown.
Second, the service provider doesn’t have to honor the counter-notification or even give the user notice of the impending takedown. The incentive to do so is to avoid liability to the uploading user for damages resulting from the takedown. But that incentive isn’t actually very incentivizing. For one thing, nearly every service provider forces its users to waive any such claims against it, in exchange for the right to (among other things) upload stuff. For another, what damages, precisely? To be honest, the main incentive to carry out the counter-notification process is as a service to one’s customers: use our service and you’ll at least get a chance to put your stuff back up.
Third, a very surprising number of uploading users are not even aware of their right to send a counter-notification, or what information must be in the counter-notification.
Fourth—and eventually most important—most service providers can take material down even without a proper DMCA takedown notice. After all, their users signed away their right to complain about it. Eventually, rights holders would start working with service providers. As much as Google and the RIAA publicly feud, they also work together quite a lot.107Just not as much as the RIAA would like and more than Google feels comfortable doing. But you know what they say about compromise.
The system is also subject to a couple of different kinds of abuse. It has been used to remove speech that someone didn’t like, or to gain an unfair business advantage.108These two examples were based on alleged trademark infringement. There is a pervasive belief that the notice-and-takedown procedure may be used for any potential legal wrong. It’s limited to copyright infringement. I’ve also heard complaints from smaller rights holders that uploading users have filed counter-notifications based on highly specious theories of “fair use.”109Which might be based on self-serving misconceptions of fair use. Uploading someone’s entire movie isn’t fair use, you know?
But from the copyright owners’ point of view, the more pertinent problem is that a takedown notice is effective only as to a specific act of infringement. If there are 20 copies of your work on a site, you’ll need to call out all 20 locations. It’s not enough to say: I own X, and all instances of X on your site are unauthorized. The theory is that each act of potential infringement should be analyzed on a case-by-case basis. One uploader might not have a license, but another might. But it makes a lot more work for the right holders, who often compare the process to playing whack-a-mole.

Isn’t That a Bit Subjective? The Limited Use of § 512(f)

Congress foresaw the possibility of abuse. It added a provision—§ 512(f)—that made you liable for sending a takedown notification or putback counter-notification that “knowingly materially misrepresents … that the material is infringing or … was removed or disabled by mistake or misidentification.” Congress even put some teeth into this provision by including attorney’s fees as an element damages, so you don’t have to be directly harmed to have a claim.
There is some tension between § 512(f) and the requirements of the takedown notification and the putback counter-notification. Those require only that the copyright owner have “a good faith belief that use of the material in the manner complained of is not authorized by the copyright owner, its agent, or the law.” (Keep that word authorized in mind. We’ll be coming back to it soon.“ And that an uploading user have only a ”good faith belief that the material was removed or disabled as a result of mistake or misidentification of the material to be removed or disabled.“110It is assumed, but by no means assured, that ”mistake“ means more than just a mistake as to authorization, but a legal mistake that would include all legal defenses, including fair use. Good faith is a term lawyers and courts understand: it means something much like subjective belief. You can act in good faith even if your actions are objectively unreasonable. Since people usually act with the subjective belief that what they’re doing is right, and people who don’t are also usually the kind of person who would be avoid leaving incriminating evidence. It is very, very difficult to overcome the ”good faith” standard.
Section 512(f) forbids making a “knowing misrepresentation” that the material is infringing. Knowing is a fraught term. It could meant that you’re required to know that the material is infringing, which is an unfairly high standard; that you must not be certain that the material is not infringing; or something in between. And by what standard are you being judged? Some acts of infringement are obvious, but many are debatable. And many of the copyright holders seeking to remove material won’t be copyright lawyers, or any type of lawyer.
In Rossi v. MPAA, the Ninth Circuit held that “knowing” had to be harmonized with “good faith.” Therefore, it is only a violation of § 512(f) if you send a takedown notification (or putback counter-notification) without even a subjective belief that the material infringes your copyright (or that the material was removed wrongfully).
There we have it. The DMCA tries to balance a desire to encourage the growth of the internet with the rights of copyright holders. It did so by taking away from rights holders one powerful but unwieldy tool—lawsuit for contributory infringement—and replacing it with a quick-and-dirty tool, the takedown notification. And it includes a provision—§ 512(f)—to encourage everyone to play nice. From my recollection of debates in 1998 and the legislative history, Congress envisioned that the takedown procedure would be quick but not thoughtless. And based on what Congress knew in 1998, that wasn’t asking too much.

Victim of its Own Success?

What Congress didn’t plan for was the immediate and exponential growth of the internet over the next few years—much of which was the result of the DMCA. In 1998, few of us had anything like “broadband”111I had a 56K modem, which felt like a huge upgrade in 1997. Just downloading music seemed like too much of a pain to be worthwhile, even with MP3 compression. Video was inconceivable. But Universities did have broadband, the thirst of graduates for broadband led to the increasing availability of broadband112And remember: U.S. broadband is very poor compared to other industrial countries’. Better compression and transmission techniques reduced download times. Streaming technology got better113Did you know the first streaming event was way back in 1995? Progressive Studios, how we miss you!, and worked better and better with commercial browsers.
The result was an exponential increase of stuff being uploaded, and stuff that rights holders had to sift through. Congress’ vision of quick-but-thoughtful became more and more untenable. Copyright law isn’t the most tractable kind of law, and soon there were more items of content there were brain cells evaluating it.114Eventually, the “solution” was for rights holders to have more direct access to hosting sites like YouTube, and to use automated programs like Audible Magic. These have rendered much of the DMCA takedown procedure irrelevant since they take place outside of its rules. Mistakes would be made. The fact that users hardly ever fought back and were generally ignorant of their rights under the DMCA did little to encourage overwhelmed rights holders to do a better job.115You could argue that the DMCA removed any incentive for service providers to come up with new and better ways to manage rights. That’s an exaggeration—there have been innovations on this issue—but not a complete one.

Let’s Go Crazy

Way back in 2007116!!!!!!!!!!, Stephanie Lenz recorded her children playing in her kitchen, while Prince’s117At the time, Prince was in a dispute with his label and wasn’t called Prince. He was instead known by an unpronounceable symbol, and everyone took to calling him “The Artist Formerly Known as Prince.” Prince is awesome. “Let’s Go Crazy” was playing on the radio. She asked her younger child what he thought of the music, and toddler stopped and started dancing (in a toddler-like way) to the music.118Because who could resist? It’s Prince, dammit![/ref] She uploaded 29 seconds of the video to YouTube. The sound quality of the music is what you’d expect—terrible.
The video is one of the most innocuous things you’ll ever see. Nevertheless, soon thereafter, Prince’s “publishing administrator”119Someone who takes care of Prince’s songwriting copyright., Universal, sent YouTube a takedown notification. Universal is, as you would expect, fairly sophisticated about copyright matters. It had an “assistant in the legal department” review YouTube videos for Prince songs. His instructions were to send takedown notifications if the music was recognizable and the focus of the video. He did not explicitly consider fair use, but it’s worth noting that his criteria would’ve been relevant to a fair use analysis. Because the video went on for more than a few moments and Ms. Lenz drew her child’s attention to the music, her video met the criteria.
YouTube duly took the video down, but it also complied with the counter-notification procedures. It notified Ms. Lenz about the takedown and the reasons for it. She obtained counsel and sent YouTube a proper counter-notification. YouTube duly put the video back.

Testy Test Case

And… that wasn’t the end of it. I wish I knew what happened next. Did Ms. Lenz demand compensation for her attorney’s fees, and did Universal foolishly decide to take a hard line? Or did Ms. Lenz just sue and force Universal to defend its actions?
Anyway, Ms. Lenz filed a lawsuit against Universal for violating § 512(f). She knew Rossi made her case difficult, so she focused on Universal’s failure to explicitly consider fair use. She argued that the video was so obviously a fair use that Universal was on a kind of inquiry notice to at least consider fair use. By failing to consider fair use, Universal could not have formed a subjective belief about infringement. After all, if you did no investigation at all—if you were making determinations based something completely arbitrary—you couldn’t be said to have any belief at all. To Ms. Lenz, Universal was liable because its investigation was woefully and fatally incomplete.
At trial, Universal put forth what seems to me a pretty persuasive argument. It argued it was not required to even consider fair use at all because it was only required to consider different types of authorization. Remember what it was that the copyright owner had to have a good faith belief about? “…that the use of the material in the manner complained of is not authorized by the copyright owner, its agent, or the law.” Ms. Lenz argued that fair use was an “authorization” by “the law.” To Universal, authorization implies something like directed permission, like licenses: “Yes, you may do that.” Fair use is more like an excuse, or, at most, passive permission: “Um, sure, I guess it’s OK if you do that.”
Comparing fair use to authorization seems kind of flabby argument. If fair use was your mother, and you were trying to find out what you’re allowed to do, and like a kid, you were pushing the boundaries, your conversation would like this:

YOU: Can I do this?
FAIR USE: Um, sure.
YOU: How about this?
FAIR USE: Ummmm, maybe?
YOU: Well, how about this?
FAIR USE: I don’t know? I don’t think so?

But how else are we to read “authorized by … the law”? How can “the law” authorize anything? If it’s legal to do something, in a way, the law is authorizing it, right?

License to Compel a License

Universal had a good argument for that. Just as you can get authorization from owners in the form of licenses, you can get authorization from the law in the form of … statutory licenses.120Also known as a “compulsory license.” That’s right: the Copyright Act has a bunch of provisions that force copyright owners to license their works to certain persons, under certain circumstances, for a certain amount of money. The best known (and easiest to understand) statutory license is § 115(a), which allows you to record and sell “covers” of other songs, so long as (a) the song has legally been recorded and sold previously, and (b) you pay the copyright owner 2 3/4 cents per work or 1/2 cent per minute.121Which is a great, great deal. Which is also why there are so many covers. For more complicated but more important statutory license, see my discussion of the Aereo case, where I talk about the statutory license that allows cable companies to retransmit signals.
This argument is persuasive because it fits within the Copyright Act’s scheme and is more consistent with “authorized” than “anything that’s not illegal.” But Ms. Lenz had a pretty strong argument in return. It is much easier to be eloquent about fair use than about, say, compulsory licenses. No authority less than Ruth Bader Ginsberg, the most “pro-copyright” Supreme Court justice, has held that fair use arises from the First Amendment right to free expression—and free expression really is something to get excited about. What’s more, if you follow Ginsberg’s logic, fair use is woven into the weft of copyright law. Those statutory licenses? They’re more like sops thrown to industry groups that might’ve made sense in 1965122The one governing musical covers seems particularly antiquated., or are kludges necessary to avoid the transaction costs of traditional licensing123I.e., without the statutory license, we wouldn’t have nice things because by the time the copyright owners agreed to a license, the potential licensee would have gone out of business. Remember it took Spotify two years to get all the necessary permissions. When Congress drafted “authorized by … the law,” could it really have been thinking only of statutory licenses and not the most important, most pervasive, most maddening affirmative defense of them all?
The trial court was more persuaded by Ms. Lenz’s argument. I might’ve gone the other way, but, again, the issue is hardly clear-cut. Universal appealed, and…

That’s Affirmative!

Well, this is where it gets a bit weird. Universal must have lost faith in its main argument, because it all but abandoned it. Instead, it tried to draw a distinction between “authorization” and “affirmative defenses.”124Quoting from Universal’s appellate brief: “Fair use is not an ‘endowment’ or ‘sanction’ of authority. It is an affirmative defense…” (emphasis in the original). With an affirmative defense, you are not liable, even if the plaintiff or prosecutor can prove their case. Some are dramatic. If you kill someone in self-defense, you are not criminally liable for homicide. Some are technical. If the plaintiff waits too long to bring a lawsuit, you are not liable, even if you did some terrible stuff.
What makes an affirmative defense an affirmative defense is who bears the burden of proof. If the defendant does, it’s an affirmative defense. In a homicide case, the defendant would need to prove that he acted in self-defense. The prosecutor doesn’t need to prove that you didn’t.
There is no question that fair use is an affirmative defense. None at all. This is so even though the Copyright Act describes fair use as “not an infringement.” The defendant must plead and prove fair use—sort of. Of the four fair-use factors, three—nature of the use, nature of the underlying work and amount and centrality of the amount taken—are pretty much self-evident. The fourth—effect on the market for the underlying work—is, practically speaking, up to the copyright owner to prove, since it has access to the relevant evidence.125If the copyright owner presents no evidence of, say, a robust licensing or distribution program, the court isn’t going to assume one for it.
This new argument is puzzling because it gets away from the actual statutory language. This is all about what’s “authorized,” remember? The distinction should be between things that are authorized and things that aren’t. Worse, the distinction between “authorization” and “affirmative defenses” just doesn’t exist. Why? Because the one thing that everyone agrees is “authorization”—the explicit license—is itself an affirmative defense. That’s right: license is an affirmative defense. The Federal Rules of Civil Procedure say so.126Software companies use this to their advantage when conducting feared “software audits.” Once they’ve established you have a certain number of copies of certain software titles, they make it quite clear that it’s 100% up to you to convince them that each copy is properly licensed. Thus, while saying that it need not consider affirmative defenses, Universal had to simultaneously admit that it did need to consider affirmative defenses.
If Universal was trying to draw a distinction between two different kinds of affirmative defenses—those that excuse behavior and those that are permitted—it didn’t make that very clear. Even if it had, the distinction has no real basis in law. If you kill someone in self-defense, no one cares if your behavior is excused, or if you are exercising a right to self-preservation. To repeat: what makes an affirmative defense an affirmative defense is who bears the burden of proof.

Fair Use Is Hard, Except All the Times When it Isn’t

Universal next argued that it shouldn’t have to consider fair use because fair use is difficult, and that’s inconsistent with the quick-and-dirty nature of the notice-and-takedown procedure. The proof might be easy, but the analysis is hard. The IP Breakdown wouldn’t ask you to play Is it Fair Use? otherwise. But, when we say fair use is difficult, we don’t really mean that fair use is always some intractable problem. What we really mean is that fair use defenses that are actually brought and adjudicated tend to be intractable. There are a great many other cases of fair use that are so easy that they don’t result in litigation. Consider:

  • A YouTube video in which someone plays a short part of a copyrighted song in order to explain some musicological concepts about the song.
  • A YouTube video in which someone plays part of two songs, to explain how one might be infringing another.
  • A YouTube video in which someone makes a “super-cut” of twelve five-to-ten second clips from well-known movies showing the same cinematography technique in different contexts.

Sorry, these are easy fair use cases. They are all for the purposes of criticism. They are all short—and no longer than necessary. None of them affect any market for the underlying work. You don’t read about cases like these because they don’t result in lawsuits; and they don’t result in lawsuits because everyone knows these are fair uses. Similarly, a YouTube video that just an unedited copy of a copyrighted motion picture is just as obviously not a fair use. True, there are some nuts who will argue anything, but for everyone who isn’t a nut, these are not difficult calls.
Universal’s argument that fair use is in all cases, at all times, a difficult determination just isn’t so. Remember that all Ms. Lenz is arguing is that her video is one of the easy cases.
Universal then argues that it can’t possibly make fair-use determinations because too much of the proof is in the user’s hands. But that’s not true at all. Three of the factors are self-evident: nature of the use, nature of the underlying work, and amount and centrality taken. The fourth—effect on market—is in Universal’s hands. Universal argued it can’t tell if a video is commercial or not, since YouTube makes money. This argument doesn’t pass the laugh test. No ads pop up in connection with the video. And—well, for goodness’ sake—just look at it. It’s clearly an amateur video uploaded because Ms. Lenz thought it was amusing.
Only then does Universal make the argument that “authorized by … the law” is actually a reference to statutory licenses. If you hadn’t read Universal’s briefs at trial (and the Court of Appeals did not), you’d think this was tacked on at the last minute.

Would Anything Have Persuaded the Ninth Circuit?

Not that any of this mattered to the Ninth Circuit because it didn’t read Universal’s brief very carefully. If it had, it wouldn’t have written: “Universal’s sole textual argument is that fair use is not ‘authorized by law’ because it is an affirmative defense that excuses otherwise infringing conduct.” As you already know, Universal had another, better argument, but it got demoted and lost in the shuffle. Actually, it suffers an even worse fate than that. Because statutory licenses are also affirmative defenses, the Ninth Circuit takes the argument as an admission that Universal must take all affirmative defenses, fair use, into account. Oh, bravo!
Worse, the Ninth Circuit doesn’t realize that Universal’s argument about affirmative defenses is a red herring and, therefore, feels compelled to take up the gauntlet. It creates a confusing distinction between two types of affirmative defenses: those that are “labeled as such due to the procedural posture of the case,” and those “that excuse[] impermissible conduct.” It then quotes from a footnote of an Eleventh Circuit case involving software, which is explicitly labeled dicta127I.e., not binding. “Although the traditional approach is to view ‘fair use’ as an affirmative defense, this writer, speaking only for himself, is of the opinion that it is better viewed as a right….” (emphasis added). that tries to draw a distinction between mere affirmative defenses and those that are really rights in disguise.
Once more with feeling: There are no distinctions to be drawn between different types of affirmative defenses. Again: if you shoot someone in self-defense, does it matter one jot whether your action is excused or if you’re exercising a right? If you re-sell a copyrighted product, does it matter that your distribution of a copyrighted item is excused, or that you’re exercising your right to “alienate” your property?128The one distinction that would matter is whether the affirmative defense is based in common law, on a statute or on the Constitution. If the latter, then even Congress can’t repeal it. Fair use is probably a Constitutional defense. But that doesn’t make it otherwise any more exalted than any other defense.

Did the Needle Even Move?

You’d think from reading the immediate analyses of the Lenz decision, copyright owners had suffered some major defeat, and it is a great victory for fair use. See here, here, here, and here129Oh, and this hysterical take. While it is an abstract victory for fair use—it’s now a right, for whatever that’s worth—it doesn’t really come at the expense of copyright holders. This is because Lenz re-affirmed Rossi. Indeed, as Sarah Jeong put it, bitterly, Lenz is “just some pretty flowers atop the grave of section 512(f).” Harsh, but not wrong, if that’s your point of view.
What that means is: rights holders must take fair use into account when sending takedown notifications, but only a little bit. This is because the process is meant to be quick-and-dirty, and your analysis really is just subjective.

To be clear, if a copyright holder ignores or neglects our unequivocal holding that it must consider fair use before sending a takedown notification, it is liable for damages under § 512(f)130Which will rarely amount to any more than a nominal amount. If, however, a copyright holder forms a subjective good faith belief the allegedly infringing material does not constitute fair use, we are in no position to dispute the copyright holders’s belief even if we would have reached the opposite conclusion.
In order to comply with the strictures of § 512(c)(3)(A)(v)131I.e., the part where say you have a good faith belief the use isn’t authorized, etc., a copyright holder’s consideration of fair use need not be searching or intensive… We are mindful of the pressing crush of voluminous infringing content that copyright holders face in a digital age….

It really isn’t asking too much, no matter what anyone says. You are only being asked to spare a thought for the rights of the uploading user. Nothing more. If this means you have to “think twice,” so be it. If thinking twice makes you reluctant to enforce your rights, I really don’t know how to help you. If you don’t think you’re qualified to make even a subjective evaluation of fair use, then learn. It won’t take long. It pays to learn the basics of the law that governs your livelihood.132This applies to everyone. You’d be surprised how often manufacturers and distributors don’t know the basics of Article 2 of the Uniform Commercial Code, even though it governs the buying and selling of goods. Enforcing rights always requires a little bit of courage—and intelligence.
You might start by questioning your own motivations. Do you wish to remove the material because you don’t like what it says? Perhaps you should reconsider fair use. Do you wish to remove the material because you don’t like the way your brand is being used? Please start over and pay attention to the part about “copyright.” Is it an amateur video in which your song is playing in the background for 29 seconds? Maybe you should just let that one slide.
Thanks for reading!

Rick Sanders

Rick is an intellectual-property litigator. He handles lawsuits, arbitrations, emergency injunctions and temporary restraining orders, opposition and cancellation proceedings, uniform dispute resolution proceedings (UDRPs), pre-litigation counseling, litigation avoidance, and other disputes, relating to copyrights, trademarks, trade secrets, domain names, technology and intellectual-property licenses, and various privacy rights. He has taught Copyright Law at Vanderbilt University Law School. He co-founded Aaron | Sanders with Tara Aaron-Stelluto in 2011.