Copyright Laws and Cases

RIAA Lawsuits

• the ISP is your school, and releasing school records is illegal.    
     (releasing names is not illegal)
  
  
• you didn't know it was against the law.
        Yes you did. Come on. But it doesn't matter.
  
  
• you already owned the tracks on CD.
  See the Gonzalez case (www.eff.org/wp/riaa-v-people-five-years-later). Almost all of Ms Gonzalez's downloads were of music she had already purchased in CD form; she found downloading to be simpler than ripping. Copyright law allows you to make a backup copy of what you bought, but there is no provision for receiving your backup copy from someone else.

There is a very limited legal doctrine of parental responsibility. Originally, the RIAA did sue parents, or made them settlement offers. More recently, after some reversals, the RIAA has been suing the minors themselves. This is a little tricky; the court must appoint an attorney, often at the RIAA's expense. Also, in Capitol_v_Foster, Deborah Foster eventually won $68,000 in legal fees from the RIAA. Foster's daughter did the downloading. (The case was brought in 2004; the RIAA dropped their suit a year later but Foster continued with her countersuit. The judge eventually ordered the award for legal costs without a full trial.)

Tanya Anderson also won a financial settlement from the RIAA, some $107,000. She was accused of downloading gangsta rap, but the facts on her disk drive didn't support that at all.

Jammie Thomas-Rasset case

Joel Tenenbaum case

Consider, though, the theory that file sharing is a violation of their copyrights, and that such individual lawsuits are the ONLY way to proceed. (Ok, there is also the idea that if the music industry had adopted streaming-on-demand earlier, then the problem would have gone away. But that's pretty speculative.)

The EFF article above, www.eff.org/wp/riaa-v-people-five-years-later, suggests that the music industry just has to adapt, but what to do is a problem. The EFF's suggestion, voluntary monthly subscriptions for the right to fileshare, seems unlikely to improve things.

• statutory damages for infringement
• rules for defendants who cannot afford an attorney
• rules of evidence

RIAA-2

• notify users of infringement for the first offense
• cut off their internet access (perhaps slowing it for a while, first)

1. file-sharers are also huge bandwidth hogs. This one was once important, but everybody streams content now, so maybe not as much today.
2. The ISP might get sued. They would probably win, but it would be an expensive hassle.
3. It's the Right Thing To Do. Knowingly cooperating with copyright infringement is wrong.
4. The RIAA might decide to sue only customers of ISPs that have not cooperated with the cutoff proposal. This is potentially a very big deal.
5. Eventually, the RIAA is likely to press for laws requiring ISPs to cooperate. Better get started now. Think about SOPA-2 (a mandatory ISP-cooperation rule was not actually part of SOPA-1).
6. The content owner might offer the ISP a cut of any settlements collected; Rightscorp apparently did this to get Cox to agree to send infringement notices to its customers. See http://torrentfreak.com/rightscorp-offered-internet-provider-a-cut-of-piracy-settlements-150525/.
7. There is a clause in the DMCA that suggests this might be mandatory (17 USC 512(i)). Somewhat surprisingly, this hasn't been extensively litigated.

(i) Conditions for Eligibility.—

See also Rightscorp below.

ISPs and copyright infringement: the Copyright Alert System

The basic model was the six strikes system: the owner of an ISP account that had been associated with infringement would receive a series of escalating warnings, with penalties. Bandwidth throttling, and possible selective traffic blocking, would be introduced on the fifth strike.

The CAS system began operation in 2013. Contesting a notice of potential infringement was complicated: it could only be done if traffic-throttling measures were to be imposed (typically after the fifth strike), the grounds for contesting the accusation were relatively narrow (Fair Use was one ground; another was that the infringement was done by unauthorized users and the subscriber could not prevent the infringement), and there was a $35 nonrefundable filing fee.

ISP customers could not protest on the grounds that they had no technical ability to identify the infringer, or that they had no legal right to monitor traffic to identify the infringer, or that they personally were not the infringer.

CAS was discontinued in January 2017, mainly because some content owners -- the MPAA in particular -- did not feel it was successful in discouraging repeat infringers. There were also problems with the fact that "alerts" were legally rather vague. On the one hand, they were clearly serious, but, on the other hand, they had no recognizable legal status.

The MPAA is now pursuing litigation against ISPs that it feels tolerate "repeat infringers". See #repeat below. The legal idea here is that ISPs are in fact required by the DMCA to cut off "repeat infringers".

The DMCA and Repeat Infringers

Because this is part of the "takedown" law in which "service providers" are clearly websites hosting content, some people initially thought this passage did not apply to consumer ISPs; that is, the "service provider" had to be at the hosting end. But this seems clearly incorrect; 17 USC 512(k)(1) says

    the term “service provider” means a provider of online services or network access, or the operator of facilities therefor....

There is no legal definition of "repeat infringer"; some suggest it means anyone who has infringed twice while others suggest that it means someone who has been legally determined to be an infringer in the past. But Cox appears not to have taken "repeat infringer" seriously at all, and "terminations" were simply account suspensions; accounts were reactivated upon customer request.

This case may mean that other ISPs start taking repeat infringers much more seriously. The Ninth Circuit ruled in 2018 that the account-termination policy didn't have to be written down anywhere (torrentfreak.com/repeat-infringer-policy-doesnt-have-to-be-spelled-out-appeals-court-rules-180324)

The RIAA model today

Prenda Law

Rightscorp

Rightscorp has gone back to attempting to use DMCA subpoenas against ISPs; courts have differed on whether this is a legitimate tactic. The "standard" tactic is to sue the individual John Doe user first, but that is expensive.

Server-based File-sharing

Fair Use

Fair Use is the right to publish or distribute copyrighted material, without permission. Generally speaking, Fair Use has to be rather limited use. There are two broad theories as to why we need Fair Use:

• Fair Use represents the minimum public access needed to make copyright law workable in situations where requiring permission is likely to have adverse effects. For example, if reviewers have to have permission to quote brief passages of a book being reviewed, the integrity of the reviewing process is undermined. This is the "limited" view.
• Fair Use represents the public's "share" of a copyright. A use qualifies as Fair Use whenever the public's rights outweigh those of the creator. This is the "expansive" view.

Legal basis for Fair Use

One of the rights accorded to the owner of copyright is the right to reproduce or to authorize others to reproduce the work in copies or phonorecords. This right is subject to certain limitations found in sections 107 through 118 of the copyright act (title 17, U.S. Code). One of the more important limitations is the doctrine of "fair use." Although fair use was not mentioned in the previous copyright law, the doctrine has developed through a substantial number of court decisions over the years. This doctrine has been codified in section 107 of the copyright law.

Here is the text of Section 107. It is short. But if you click on the Notes tab, you get the explanatory House Report 94-1476, which is quite lengthly. This Report is the kind of material the late Justice Scalia felt should not be used by the courts when interpreting the law. Notice also the rather extensive discussion on the photocopying of material for classroom use.

1. the purpose and character of the use, including whether such use is of commercial nature or is for nonprofit educational purposes;
2. the nature of the copyrighted work;
3. amount and substantiality of the portion used in relation to the copyrighted work as a whole; and
4. the effect of the use upon the potential market for or value of the copyrighted work.

• South Park (almost any episode)
• Weird Al
• www.xkcd.com/c78.html
• Bored of the Rings
• 2 Live Crew and the Campbell case
  

The Ford quotation also led to the cancellation of a contract to serialize portions of the book in a sequence of magazine articles. That reflects factor 4.

Music sampling, in the sense of 1-2 second snips used in another work, might be Fair Use. 10-20 seconds is a lot longer.

Factor 4 is the big one. See Sony v Universal. A tricky problem with Factor 4, however, is that while there might not be a market now for the use in question, such a market could potentially develop. That is, a market for music sampling rights might develop (has developed!) if sampling were not claimed as fair use. A market for prerecorded television shows has definitely developed. Later we'll consider a case in which the plaintiff claimed that they were considering marketing thumbnail images, and thus images.google.com's "republication" of thumbnail images was not Fair Use.

If copyright were an ownership right, rather than a utilitarian attempt to create an incentive for content creators, there would be no room for Fair Use.

Judge Pierre Leval and Transformative Use

Leval wrote a Harvard Law Review paper in 1990 on Fair Use; his goal was to try to identify some fundamental principles. Leval was very clear that the justification of copyright was entirely utilitarian. One of the principles Leval identified was that use should be transformative.

Leval thought that, to be fair use, the use must be of a character that serves the copyright objective of stimulating productive thought and public instruction without excessively diminishing the incentives for creativity. For Factor 1, the "purpose and character of the use", the most important question is whether the use will serve to "fulfill the objective of copyright law to stimulate creativity for public illumination".

This was the original meaning of transformative use; there is no physical or computational transformation involved.

Leval spends a considerable amount of time discussing biographies, and whether the authors of biographies can claim fair use when quoting the letters of their subjects [Leval heard at least two significant cases in this area]. This is a tricky question, as the letters often have value when published alone. On the other hand, Leval asks "can ideas be correctly reported, discussed, or challenged if the commentator is obliged to express the idea in her own different words?" Or how about this observation:

If a biographer wished to show that her subject was cruel, jealous, vain, or crazy, can we seriously contend she should be limited to giving the reader those adjectives, while withholding the words that support the conclusion?

One issue is whether not-yet-published work can be excerpted by others as Fair Use. Some courts have flatly said no. Leval argues that this should apply to work intended for eventual publication; authors should basically have considerable control in how their work is first published. But a lot of writing -- letters, for example -- is not written with eventual publication in mind. Again, the ultimate rule, Leval argues, should be whether the fair use will interfere with the incentive to create works in the first place. Publication of early drafts can do exactly that; publication of letters, not so much. Leval raises the case above of publication of 300 words from Gerald Ford's autobiography: this was pre-publication use. However, even if it were post-publication use, it might still not be fair.

When we look at music sampling, we will see that one argument against Fair Use is that, while samplers may do no harm to the market for the original recording, they always do harm to the market for sampling rights. Leval argues against that:

By definition every fair use involves some loss of royalty revenue because the secondary user has not paid royalties. Therefore, if an insubstantial loss of revenue turned the fourth factor in favor of the copyright holder, this factor would never weigh in favor of the secondary user.

Leval also points out that "an adverse criticism impairs a book's market", but nobody thinks that is the kind of "effect on the market" that is meant here. The use in question must serve as a substitute, not merely discourage sales of the original.

1. The primary use of the defendant's system was for infringement
2. Aimster was providing a service rather than a piece of hardware
3. There was an ongoing relationship between Aimster and its customers
4. Aimster's service enabled not only home copying, but distribution
5. The system was specifically designed to enable infringement
   

Criminal copyright violations

17 U.S.C. § 101
§ 101. Definitions

Napster

Music Sampling

• Ulrich Schnauss
• Guns N' Roses (analog copy without ads here)
  

• rap "remixes" that use someone else's entire song
• musical "references" to other songs, meant as an homage
• "mosaic" sound made up of many tiny samples

Note that GnR and Schnauss do not appear to fall into any of these categories. Some of Gillis' work seems to fall into the first category; some into the third.

The Bridgeport Case

The industry line here is that any use of copyrighted material requires permission; this gives the rights-holder the opportunity to set a fee limited only by the law of supply and demand.

In 2015 Robin Thicke and Pharrell Williams lost their Ninth Circuit appeal of copyright infringement case from the estate of Marvin Gaye, claiming infringement by Thicke and Williams' song "Blurred Lines". The song was apparently intended as an homage to the "feel" or "groove" of Gaye's song "Got to Give It Up", and the "groove" of that 1970's era. It is quite plausible that Thicke and Williams might have prevailed had they introduced a Fair Use claim, but they did not -- again because they failed to raise it at the district-court level. So we still don't have a good case that settles Fair Use one way or the other. The Thicke/Williams song did not involve sampling, however.

Look at the recent case Cariou v Prince, below, for an important and relevant fair-use case in the visual arts.

The Ciccone Case

In 2016 the Ninth Circuit decided VMG SalSoul v [Madonna] Ciccone, in which they ruled that de minimis sampling was definitely a thing, and Ciccone's horn samples in Vogue -- sampled from Love Break -- were not infringing because the use was too small. The samples were around a quarter-second in length. The song Vogue was released in 1990.

This does directly contradict the Bridgeport decision. The Ninth Circuit argued that Congress had not eliminated the de minimis defense in the context of sound recordings.

Once again, Fair Use was not considered.

An important consideration for the court was that "a general audience would not recognize the brief snippet in Vogue as originating from Love Break." This might be a part of the court's understanding of de minimis: a sample too small to be recognized is a sample too small to matter. That said, it does suggest that de minimis sampling can't apply when the sampling is an homage, because such samples are supposed to be recognizable.

The horn sounds of the Ciccone case were quite a bit shorter than the sample of the Bridgeport case, but, in general, from the reasoning of the Ciccone case it seems like they would have decided Bridgeport similarly.

Because the Sixth and Ninth Circuits are now in conflict, the Supreme Court is more likely to take up a case regarding de minimis sampling. But, to date, they have not.

The opinion is here: cdn.ca9.uscourts.gov/datastore/opinions/2016/06/02/13-57104.pdf.

Theft! A History of Music

The Duke Law graphic novel, Theft! A History of Music, addresses the Bridgeport case starting at https://web.law.duke.edu/musiccomic/read/large/173/. The book takes a rather strong Fair Use position on sampling. It is definitely worth looking at.

The Amen Break

They don't make millions from the drum break, however. And we don't actually know what Coleman himself thought of all the times he was sampled.

Drake vs Estate of Jimmy Smith: A Fair Use victory? Sort of.

The rap musician Drake used a sample from a recording by Jimmy Smith on his track Nothing Was the Same / Pound Cake. The judge ruled in 2017 this was Fair Use. Sort of; Drake in fact had obtained a license to sample the performance, though not the "composition". See law.justia.com/cases/federal/district-courts/new-york/nysdce/1:2014cv02703/425904/135/.

Smith's track was spoken, not sung, which means that the existence of a "composition" is iffy. It was the final track of his jazz CD. Smith originally said "Jazz is the only real music that’s gonna last.". Drake edited the words so they said "Only real music is gonna last". Here is Smith's original; underlined passages were used by Drake (below).

Good God Almighty, like back in the old days.
You know, years ago they had the A&R men to tell you what to play, how to play it and you know whether it’s disco rock, but we just told Bruce that we want a straight edge jazz so we got the fellas together Grady Tate, Ron Carter, George Benson, Stanley Turrentine.
Stanley was coming off a cool jazz festival, Ron was coming off a cool jazz festival.  And we just went in the studio and we did it.
We had the champagne in the studio, of course, you know, compliments of the company and we just laid back and did it.
Also, Grady Tate's wife brought us down some home cooked chicken and we just laid back and we was chomping on chicken and having a ball.
Jazz is the only real music that's gonna last.  All that other bullshit is here today and gone tomorrow.  But jazz was, is and always will be. 
We may not do this sort of recording again, I may not get with the fellas again.  George, Ron, Grady Tate, Stanley Turrentine.
So we hope you enjoy listening to this album half as much as we enjoyed playing it for you.
Because we had a ball.

The full edited passage Drake used, 35 seconds long, is here:

Good God Almighty, like back in the old days.
You know, years ago they had the A&R men to tell you what to play, how to play it and you know whether it’s disco rock, but we just went in the studio and we did it.
We had champagne in the studio, of course, you know, compliments of the company, and we just laid back and did it.
So we hope you enjoy listening to this album half as much as we enjoyed playing it for you. Because we had a ball.
Only real music is gonna last, all that other bullshit is here today and gone tomorrow.

Smith said only jazz was going to last. Drake disagrees. (It is not clear of the change from "that's" to "is" in the last line represented the patching in of a copied "is" from somewhere else, or if it's just a variant of transcription, with the underlying sound unchanged.)

Drake's team had obtained a license to sample from Smith's recording. But they did not have a license to sample from Smith's composition. The only argument Smith's estate had was that Drake used a sufficiently large amount of Smith's track that they needed a composition license too.

One way to look at this is to argue that the situation would be the same if Drake had simply copied Smith's words in print. As a quotation. Who would care about that?

Judge William Pauley gave a careful Fair-Use four-factor analysis, ultimately ruling that Drake's use was transformative. On the purpose of the use factor (the first factor), Drake's legal team had argued that Drake's track (and Drake's message) was quite different:

1. Drake changed the meaning considerably, by dropping the reference to "only jazz".
2. Drake dropped all references to Jimmy Smith's own musical work
3. Drake added background music and generally completely changed the "feel" of the piece.

The judge regarded points 2 and 3 as making Drake's track a derivative work, but felt that point 1 rose to the level of "transformative".

The judge felt that factor 2, "the nature of the work", was slightly in favor of Jimmy Smith.

For factor 3, the amount of the work used, the judge cited Cariou v Prince (remember essentially all of Cariou's original photo was used).

For factor 4, the effect on the market for the original, Judge Pauley writes "There is no evidence in the record to suggest that Pound Cake usurps any potential market for JSR or its derivatives". This is, for sampling (and especially for small-scale sampling), almost always the case. Judge Pauley did note that Smith's work "targets a sharply different primary market" than Drake's. That probably should not be relevant to a factor-4 conclusion, but there it is.

Transformative Use

"Transformative use" is use of copyrighted material where the "purpose .. of the use" (factor 1) is wholly different from the purpose of the original. Typically it may be important that the new use offer something to the public that was otherwise unavailable.

Parodies are often considered transformative use. Note, however, that permission to parody is often refused, so many parodies might not exist if it were not for Fair Use.

Another example comes from Diebold v Online Policy Group, and some Swarthmore students. The students and OPG were originally suing Diebold; the students had posted some internal Diebold memos about flaws in electronic voting machines, and Diebold was wildly filing DMCA takedown notices. The students and OPG, and the EFF, felt these were an abuse of the DMCA process.

From the judge's opinion:

Search engines often (and often successfully) raise the transformative-use defense when the alleged infringement comes up in the course of making search available. For example, Google caches whole pages of other sites, and also makes entire images available on images.google.com. The transformative-use argument is based on the idea that Google is not publishing articles or images to compete directly, but rather to provide its search service to the public. See the Perfect 10 case below.

A claim of transformative use might be raised for music sampling, especially if the sampled sound is digitally modified, but generally speaking it is the intent that matters when use is transformative. The Online Policy Group's purpose in publishing Diebold's internal memos about flaws in their voting machines was to further democracy, not to make competing voting machines.

When Google copies articles (and, as we shall see below, photographs) as part of maintaining its search engines, the purpose is to enable search, not to compete as a publisher. (One can take a cynical view that Google does indeed try to compete as a publisher, but the idea is still that search brings to the public something that was fundamentally absent before.) Baase 4e §4.3 / 5e §4.4 is devoted to this topic of search engines and copyright.

Is music sampling a transformative use? Is it transformative in some cases? If so, what cases?

The Kelly and Perfect 10 cases, next, address this doctrine of "transformative". We'll return to the Diebold case under "Free Speech"

Kelly and Perfect 10

• Purpose and Character:     use is transformative
• Nature of work:                 creative work on internet; "slightly in favor of Kelly"
• Amount & Substantiality: irrelevant; whole image must be copied
• Effect on market:              The court found no harm to Kelly's market; in fact, by helping people find Kelly's images they might help him. Use of thumbnails weighed heavily here: they aren't nearly as attractive as originals.

Ninth Circuit

1. Google DMCA defense
2. P10's "display right" and "distribution right" are at issue.
3. Server test: whose server are the images really on?]

Cariou v Prince

This case about photography was a rather substantial win for Fair Use, and in some ways a surprising one. Richard Prince used photographs by Patrick Cariou in what might be called a collage, but which might also be called trivial editing. Some of Prince's work can be seen here: http://www.gagosian.com/exhibitions/richard-prince--may-08-2014/exhibition-images. Here are some archived images:

• Graduation
• James Brown Disco Ball
• Back to the Garden
• Canal Zone

But traditionally "transformative use" means use for a different purpose, not just a different aesthetic.

The court rejected the argument that Prince's work had to be parody to be protected:

The court also looked at the commercial/noncommercial distinction. Prince was selling his work; that makes it "commercial". However, the court noted "nearly all of the illustrative uses listed in the preamble paragraph of § 107 [the Fair Use law], including news reporting, comment, criticism, teaching, scholarship, and research... are generally conducted for profit." So is music sampling, but it is also "artistic".

• It must fit into two 15-minute segments (eg YouTube segments)
• They can't use the name Star Trek in the title
• They can't earn advertising revenue from online showings

Given that the original plan was for a ~2-hour movie, 30 minutes is something of a letdown.

Search Engines generally

Sony and Reverse engineering

Sega Enterprises v Accolade, Ninth Circuit, 1992: Accolade made copies of the Sega ROM and reverse-engineered it, so that they could distribute their own games for the Sega Genesis, without paying licensing rights to Sega. This licensing ran $10-$15 per game cartridge manufactured. Licensing also would prohibit Accolade from releasing their games on other consoles! Accolade used the reverse-engineered ROM to figure out how to package their own games so as to bypass the Sega licensing check; as I understand it, they did not distribute Sega's ROM outside of their programmers' offices.

At a later point, the Sega system would only allow games that had a password at a specific point in memory, and then would display the Sega trademark. They tried to claim that causing the trademark display was itself illegal, under trademark law, regardless of copyright. Sega also tried to claim that copying the password file -- about 25 bytes -- was infringement. Accolade argued for Fair Use.

Accolade ultimately won the case. Fair Use was only part of it; there were also trademark issues, Sega's attempt to copyright their whole DRM scheme, and anti-trust issues.

Sega argued that all four Fair Use factors were in their favor. In particular, Accolade copied the entire ROM, and  The Ninth Circuit dismissed much of this, pointing out that copies of Sega's program were not being distributed at all, and were not even being used for infringing game play within Accolade. The copies were made only to create new games; Accolade's only "harm" to Sega was as competitor. The Ninth Circuit did cite the Sony decision regarding the possibility that an entire copy might in some cases be fair use, and that "[w]hen technological change has rendered its literal terms ambiguous, the Copyright Act must be construed in light of this basic purpose [to stimulate artistic creativity for the general public good]" (in turn quoting an earlier copyright case).

A major part of the decision was that you cannot copyright software function. You can only protect it via patent or trade secrecy.

Eventually, after Sega lost the case, Accolade did become a Sega licensee, though perhaps there was a special deal.

Sony Computer v Connectix, 2000: Connectix copied Sony BIOS and reengineered it so that Sony Playstation games could be played on a computer. The Ninth Circuit ruled 3-0 that the copying done by Connectix was Fair Use. Under the "nature of the use", the court was influenced by the idea that there was no way to analyze the functionality of the ROM without copying. The court also recognized that the market for Sony consoles might be affected, but not the market for any of the games played on those consoles.

Sony ended up buying Connectix out. Don't think "Connectix lost", think "nice golden parachute". Or that we lost.

A note on the 2010 Sony hack

Are individuals and groups such as Hotz and Anonymous criminals? Is Sony's response heavy-handed? Or is it a legitimate effort to protect the intellectual-property rights of third-party developers?

Google v Oracle

Sun Microsystems developed Java. Oracle bought Sun in 2010 and became the owner.

Google released Android with the standard Java API. Google wrote the implementations of the API functions, but the API itself came from Oracle. This is essentially the function prototypes:

    boolean Java.util.HashMap.containsKey(Object key)
    void Java.util.HashMap.put(K key, V value)

Oracle sued, claiming they owned even the API.

At the district-court level, Judge Alsup ruled in Google's favor, due to Fair Use. He was one of few federal judges to have experience developing software.

Normally the case would have been appealed to the Ninth Circuit, which likely would have been more disposed to Google. But Oracle got a patent claim included in the appeal, and so it went to the Federal Circuit, which is rather more IP-owner-friendly. Oracle won.

But the case bounced around a while and finally the Supreme Court heard it. They ruled for Google in October 2020. They ruled in favor of Google's Fair Use defense, and so explicitly did not consider whether APIs are copyrightable at all. Here is the Court's argument:

1. Nature of the Work: favors Fair Use. The Court was influenced by the fact that this was about interfaces.
2. Purpose and Character of the use: The Court found Google's use was transformative, in that they were simply moving Java to a different platform. It's not completely clear this is transformative in the sense of being a different purpose for the use, like the Diebold documents leak; Google is after all simply trying to encourage developers to write software for Android. However, Google's purpose was arguably to port Java to a new platform so that others (eg app developers) could write programs for that platform.
   The court also did say that "we have used the word “transformative” to describe a copying use that adds something new and important." That is in fact a different standard than the "different purpose" standard.
3. Amount and substantiality: Google copied 11,500 lines of code. This is the entire API header set, but only one part in 250 of all the code. The Court felt the latter comparison was more appropriate.
4. Effect on the market: the Android platform is not a substitute for Java SE. (Maybe the IcedTea Java clone would be?). Also, the court cited evidence that Oracle "would benefit from the reimplementation of its interface into a different market."

Hmm: it seems to have moved to https://www.youtube.com/watch?v=cZiJQnIMV_s. Ignore the opening text ad.

O'Reilly is actually talking about the person who downloaded (and publicized) Sarah Palin's email, during the 2008 election. However, many of O'Reilly's points apply to how the term "intellectual property" can be a misleading concept.

The perpetrator of the Palin hack was eventually identified as David Kernell, and was later convicted and sentenced to 366 days in jail.

Bill's correspondent is attorney Megyn Kelly. Kelly acknowledges that it is indeed a "federal offense to access email without authorization", but goes on to say that the web site is probably ok [~2:00 minute mark]. O'Reilly responds with "they're trafficking in stolen merchandise" and compares it to if "you steal somebody's car".

The relevant Supreme Court case is Bartnicki v Vopper, which established that one has a first-amendment right to publish illegally recorded conversations, provided you were not a participant in the illegal act.

At 3:13, he says there is "no difference between taking a person's letter out of the mailbox and taking someones email off their internet site".

Is there a difference?

• Theft is theft. (Is this a deontological sentiment?)
  
• Movies cost a lot
• music downloading is as bad as credit card theft
• Everyone has to play by infringement rules
• distributing a DVD is no different from stealing newspapers
• THEFT IS THEFT
• [creators are entitled to] FULL RIGHTS OF OWNERSHIP
• "Today's Internet pirates try to hide behind some contrived New Age arguments of cyberspace"

• What if anything was iCrave.com doing wrong? Apparently they were simply streaming broadcast TV to the internet in real time; who loses? They were wrapping the content in banner ads.
  
• How closely is file-sharing related to credit-card fraud and endangerment of children?
  

Can the FILM industry survive on the napster model?

• What is the FAIR amount of money to pay for something?
• Can people be EXPLOITED by receiving too low an income?
• Is HARM to other people ever justified, aside from ECONOMIC HARM?
• Are there limits to justifable ECONOMIC harm?

Napsterized Video Games

Film Filesharing

Broadcast TV Sharing

During oral arguments on April 22, the Court seemed very concerned that a verdict in ABC's favor could also undermine the broader concept of cloud-based content storage. After all, Aereo's basic business model is to allow you to lease a small antenna in New York City and record your favorite broadcast-TV shows on your own cloud space. You can then watch them at your leisure. Is there some way in which this is different from storing your purchased content in the cloud?

How can the Court separate the leased-antenna concept from the cloud-storage concept?

Justice Roberts said "I'm just saying your technological model is based solely on circumventing legal prohibitions that you don't want to comply with". Though he immediately added, "which is fine. I mean that's -- you know, lawyers do that".

Also Justice Sotomayor owns a Roku.

Digital Restrictions Management

1. encrypted media files, with multiple possible decryption keys
2. per-file, per-user licenses, which amount to the encrypted decryption key for a given file
3. player software (the DRM agent) that can use some master decryption to decrypt the per-file decryption key and then decrypt the licensed file. The DRM agent respects the content owner's rights by not allowing the user to save or otherwise do anything with the decrypted stream other than play it.

General copyright law rules

• the work is sold
• the work is a Work For Hire

Laws (highlights only):

1972: Sound recordings created in 1972 and after were brought under Copyright. Coverage was not retroactive, but state copyright coverage is allowed to remain in effect until 2067. Pre-1972 recordings are therefore never in the public domain unless the copyright holder has placed them there.

(The 2018 Hatch-Goodlatte Music Modernization Act updated this to bring pre-1972 songs under Federal copyright, and to create an expiration date. Recordings made prior to 1923 will have their copyrights expire on January 1, 2022. Later recordings will similarly get ~100 years of copyright protection.)

• indecency v obscenity and the Internet
  
• Section 230

• anticircumvention prohibition: it is illegal to help someone in any way to circumvent copy protection
  
• safe-harbor / takedown

Some Famous Copyright Cases

Copyright and traditional music

1985, Dowling v United States, 473 U.S. 207

1991,  Feist Publications v Rural Telephone Service

Paragraph 8:

MGM v Grokster, 2005

DMCA, 1999

OCILLA

DMCA Contains OCILLA: Online Copyright Infringement Liability Limitation Act. Generally, this is 17 USC Section 512 (c)"Information Residing on Systems or Networks At Direction of Users". A few later sections are also involved, eg for the "putback" provision.

• description of infringing material
• good-faith claim that use is not legal
• sworn statement requestor is authorized to act by copyright holder

• description of material
• good-faith claim use is legal, subject to perjury
• acknowledgement of court jurisdiction

Why do you think blackboard is so popular? Hint: not because it's easy to use.

There's a part of OCILLA here that has, historically, not been widely understood, which is the following:

(i) Conditions for Eligibility.—

This has been used to argue that ISPs are "service providers" and so are liable unless they have a policy for dealing with repeat infringers. The problem here is that all the "limitations on liability established by this section" have essentially nothing to do with ISPs; they apply only to content-hosting sites. But some ISPs do also host content.

Mail & Guardian

The Mail and Guardian, a South African weekly newspaper, published an article on June 21, 2019, which suggested that Njock Ajuk Eyong was engaged in some form of fraudulent or corrupt business. On Sept 10, the M&G received a notice from their online hosting provider, Linode.com, that a takedown notice had been received about the article, and that M&G had 96 hours to remove it. When the 96 hours elapsed, Linode shut their entire site down, despite M&G's furious arguments. M&G said "[o]ur questioning the veracity of the complaint did not seem to make any difference". This is true; under US law it does not make a difference. M&G reluctantly removed the article, and service was restored.

The takedown notice was ostensibly from a blogger who claimed to have published the same content on June 3; the blogger's only date verification was the date entered into the blog. See mg.co.za/article/2019-09-15-censored.

What could M&G have done? Two things, I suppose:

1. Modify the content and repost the article. They did that, in effect, by reposting the article to Twitter, but the takedown notice sent to Linode applied only to the original article. So they should have been able to repost a modified version on their Linode site as well.

2. File a DMCA putback request. But this means the article is offline for at least 10 days, which can be critical in the news cycle. It also means that M&G would be agreeing to the jurisdiction of the US courts; while this is sometimes a consideration for non-US publishers, it would not likely have been an issue here. (M&G may have done this; they may also not have known it was possible.)

17 USC 512(f) contains a clause making it illegal to file a false takedown notice. So the M&G can in principle pursue the blogger. But it is likely the blogger simply does not exist.

It does seem that Linode could have been more helpful.

Cases related to DMCA/OCILLA

• Selections would be at Warner's discretion
• Warner would control the advertising, and receive all revenues
• They would be played through a special player provided by Warner [malware alert!]

IO Group v Veoh Networks

Viacom v YouTube

Discussion

Viacom v Google District Court ruling, June 2010

(c) Information Residing on Systems or Networks at Direction of Users.:

(1) In general. : A service provider shall not be liable for monetary relief, or, except as provided in subsection (j), for injunctive or other equitable relief, for infringement of copyright by reason of the storage at the direction of a user of material that resides on a system or network controlled or operated by or for the service provider, if the service provider -

(A)(i) does not have actual knowledge that the material or an activity using the material on the system or network is infringing; [and]

     (ii) in the absence of such actual knowledge, is not aware of facts or circumstances from which infringing activity is apparent; or

     (iii) upon obtaining such knowledge or awareness, acts expeditiously to remove, or disable access to, the material;

(B) does not receive a financial benefit directly attributable to the infringing activity, in a case in which the service provider has the right and ability to control such activity; and

(C) upon notification of claimed infringement ... responds expeditiously to remove... the material that is claimed to be infringing [the takedown]

Appellate Court 2012

Willful Blindness

Back to the District Court 2013

Settlement!

Judge Stanton's latest decision was again appealed, but then on March 18, 2014, the parties settled (as a flock of flying pigs migrated northwards?) Industry rumors suggest that Google paid no damages, but did let Viacom in on the advertising revenue. In other words, Viacom caved completely.

Content ID

One of Google's early responses to claims of copyright infringement on YouTube was to develop the Content ID system, which automatically attempts to detect whether uploaded content has been copyrighted by someone else. If a match is detected, then the copyright owner can block the content, or ask for the advertising revenue.

Content ID has been quite effective at reducing wholescale unauthorized uploads. Most objections to it are that it is too strict. For example, sometimes background music trips the Content ID system.

Content ID has been particularly pernicious for those wanting to make videos of game walkthroughs; the game vendor usually claims copyright over even brief snippets of game play. But game walkthroughs are a legitimate thing.

There have been several clever workarounds to Content ID. One early strategy was to display video content within an inner TV screen, where the TV is part of the video.

A recent approach has been to include snippets of other videos -- perhaps videos of games of other manufacturers. If the Content ID system sees multiple matches, it gets sufficiently confused that it ends up not allowing any one owner-claimant to take over the video. See, for example, kotaku.com/game-critic-uses-brilliant-workaround-for-youtubes-copy-1773452452.

A Posner-style view of Viacom v YouTube

Discussion

Tiffany v Ebay

UMG v Veoh

Copyright holders know precisely  what materials they own, and are thus better able to efficiently identify infringing copies than service providers like Veoh, who cannot readily ascertain what material is copyrighted and what is not. 

If this notification had come from a third party, such as a Veoh user, rather than from a copyright holder, it might meet the red flag test because it specified particular infringing material.

Server-based Filesharing

• Not being a real cloud-storage site, because continued storage depended on other users downloading the content.
• Download users often paid for "premium" bandwidth or search capability.
• Their business model was based on advertising, and ads were shown only to downloaders
• MegaUpload was in fact successful at blocking child-abuse content, suggesting that they in fact could have blocked infringing content.
• Users who uploaded infringing content never had their accounts terminated.
• Users who uploaded popular content received rewards, either in cash or in increased bandwidth for downloads
• When takedown notices were received, only the link to the file was removed. The file itself would remain, available via other links. There are legitimate reasons for this; for example, the other links may represent personal, non-shared copies of the content in question.
• MegaUpload may have been wilfully slow at handling takedown notices, at least at some times. There does seem to be agreement that, eventually, they handled most takedowns on a timely basis. It is possible that the "slowness" issue refers to the fact that MegaUpload removed links rather than files.

For MegaUpload's purported side to this, see www.hangthebankers.com/10-facts-about-the-megaupload-scandal.

• they might allow seizure of domain names without due process
• they might gut the DMCA user-posted-content safe harbor

Either of these, while effective for taking down filesharing servers, would radically change the Internet as we know it.

SOPA and PIPA

• https://www.techdirt.com/articles/20141006/02561228743/5000-domains-seized-based-sealed-court-filing-confused-domain-owners-have-no-idea-why.shtml
• http://www.internetcommerce.org/undue-process/
• http://blog.ericgoldman.org/archives/2011/11/court_oks_priva.htm

Amazon

ASCAP

• selection of lower-cost (eg older) works, in order to save money
• proving that a significant fraction of the music played was non-ASCAP, in order to save money.

Sita Sings the Blues

Kindle case

• Book publishers have already agreed to kindle distribution
• the synthetic voice has no inflection or emotion
• The synthetic voice does not constitute a "performance"
  
• No copy is made [is this a legitimate argument?]
• Amazon is a leading seller of audiobooks; arguably they don't see a negative effect on the market.
  
• what about reading for the blind?
• what about conventional-text-to-braille scanners?
• existing audiobook formats (CDs) are unsatisfactory
• use is transformative
  

• The kindle infringes on the right to create audio recordings of books
• creation of a spoken "performance" is not a right that was granted by purchasing the text.
  
• publishers thought this was a text-only deal
• people who buy e-books to listen to while driving now may not buy the audio book.

Authors Guild v Google

The Google Books Project began in 2004 as a way to support searching within books (some have suggested that Google wanted to prevent Amazon from becoming the default book-searching site). Google, with the cooperation of a consortium of university libraries, began scanning millions of books into its database; the total eventually reached 30 million. Users of books.google.com could search this database for words and phrases.

As many of these titles were still under copyright, the Authors Guild sued. Google reached a tentative settlement agreement with the Authors Guild in 2008; this was amended in 2009. This agreement allowed Google to sell digital editions of the books, with compensation to the authors. US District Court judge Denny Chin, Southern District NY, ruled in March 2011 against this agreement, arguing at the time that it was unfair to authors. Chin had by that time been appointed to the Second Circuit, but was still finishing up with some of his previous District Court cases.

Librarians were often against the draft agreement because of Google's lack of respect for the principle of patron privacy, and out of fears that library rates for access to the Google Library would be excessive.

The parties went back to negotiation. Google began an argument before the court that their scanning amounted to Fair Use, and so they did not need authors' permissions at all.

Two and a half years later, in November 2013, an odd thing happened. Judge Chin now ruled that Google's book-scanning project was indeed Fair Use. No author permission was needed.

He wrote

In my view, Google Books provides significant public benefits. It advances the progress of the arts and sciences, while maintaining respectful consideration for the rights of authors and other creative individuals, and without adversely impacting the rights of copyright holders.

Overall, the decision is a straightforward Fair Use analysis. The fact that Google copied the entire book counts "slightly" against it, but it does not display the entire book to users, and the transformative nature of being able to provide the search feature counted heavily in Google's favor. The fact that search can lead new readers to a book appears to count even more heavily. (Recall that Justice Blackmun wrote in his Sony dissent "Even a showing that the infringement has resulted in a net benefit to the copyright holder will not suffice".)

Sometimes Google's search is considered "commercial", because Google profits from the associated advertising. Google stopped advertising on its "About the Book" pages, however, possibly with this issue in mind.

Google divides each book page into eight snippets. When a phrase is found, Google usually displays the snippet containing the phrase and two others (probably the preceding and following snippets). But some snippets are never displayed, and some entire pages are never displayed, making it impossible to obtain an entire book this way. One can obtain a large majority of it, though.

For some books, such as reference works and poetry, readers are often content with just a few pages. The proposed (and now moot) settlement had mechanisms in place to protect against this; the recent ruling appears to make no mention of it.

Why did Chin change his mind?

After the breakdown of the initial agreement, the Authors Guild upped the ante by pressing their parallel case against the Universities who had partnered with Google. This case was Authors Guild v Hathitrust. In October 2012, District Court judge Harold Baer (also of the Southern District NY, but a different judge was assigned because it was theoretically an unrelated case) ruled that the book scanning was Fair Use.

Baer gave a detailed Fair Use analysis in his ruling, which Chin no doubt read closely.

Baer was particularly moved by how the Google Books project offered enormous benefits to the blind. In fact Section 107 of the Copyright Act, defining Fair Use, gives as an example the creation of copies for the blind.

In an eloquent oral argument ... Defendant Intervenors spelled out where blind scholars stood before digitalization: "Prior to the development of accessible digital books, the blind could access print materials only if the materials were converted to braille or if they were read by a human reader, either live or recorded.". Absent a program like the [Google Books project], print-disabled students accessed course materials through a university's disability student services office, but most universities are able to provide only reading that was actually required.

I have heard a story that Judge Baer was so sensitive to the needs of the blind because he himself is deaf, though I have not been able to confirm that. But in any event Baer's decision made it abundantly clear that the actions of Hathitrust were Fair Use. And this is directly applicable to the Google case, which was otherwise muddied by questions of whether Google's use was "commercial", and details of how digital sales were to be handled.

The Hathitrust defendants may have been more sympathetic than Google: they are unarguably noncommercial.

Did the Authors Guild commit a strategic blunder by pursuing its case against the Hathitrust?

Did we somehow all lose after all?

The case firmly settled that book scanning was in the public interest for search, and thus was fair use.

But the other part of the original agreement was that it would have allowed Google to sell digital copies of books, with payment to the author (and probably to the publisher), if the author was alive or if copyright information could be found. Google would give away access to public-domain books. As theatlantic.com/technology/archive/2017/04/the-tragedy-of-google-books/523320 put it:

You were going to get one-click access to the full text of nearly every book that’s ever been published. Books still in print you’d have to pay for, but everything else—a collection slated to grow larger than the holdings at the Library of Congress, Harvard, the University of Michigan, at any of the great national libraries of Europe—would have been available for free at terminals that were going to be placed in every local library that wanted one.

Fair Use isn't going to allow this, however; just the search. The proposed agreement would have settled how to make accessible both orphan works -- books that had no known copyright status -- and out-of-print works, which were still under copyright but not actively available. As things stand now, no one can legally distribute works in either category; they might as well have been burned. Which is why the Atlantic article above is titled "torching the modern-day library of Alexandria".

Congress could act, of course, but won't. And, as the Google agreement appears to be completely dead, our best chance at solving the access dilemma appears to be lost.

Google itself has largely abandoned the book project, except for search.