Friday, December 16, 2011

SOPA on a Rope?

There has been an immense amount of talk lately regarding a bill currently making its way through US Congress known as the Stop Online Piracy Act (or SOPA).  SOPA has been fashioned as an effective response to rampant online piracy of copyrighted content by granting the Attorney General rather broad and sweeping powers to attack infringing websites. The law also gives individual rights holders a new procedure complete with remedes they’ve never had access to before.

The bill is a companion to the equally controversial PROTECT IP act, also being debated by congress. These two bills signify a sharp change in the American copyright landscape, specifically as it pertains to the Internet.

In addition to granting the Attorney-General the power to obtain a court order against a site that is either “committing of facilitating online privacy”, he may also order third party payment providers (such as Visa or PayPal) and online ad networks to stop supporting and dealing with the site in question.

photo by mikeleeorg
Furthermore, the Attorney-General may order the de-indexing of allegedly infringing sites from search engines.  Many have likened this rather extreme measure to the Internet censorship protocols enacted by the Chinese government (known coloquially as ‘The Great Firewall”).  While this may be a slightly extreme appraisal, this measure in particular brings up significant free speech issues.

As mentioned above, not only the Attorney-General but regular rights holders have been given powerful enforcement tools.  If a rights holder believes a site is infringing their copyright, they must undertake a two step process: First, they must contact payment providers and ad networks doing business with the site in question.  Upon forwarding them a compliant, they are to cease their service in relation to the website and forward the complaint to the owner of the domain. At this point, the site owner must reply and explain how their activities are not infringing copyright (eg. the material is legally licenced, fair use etc.).  if they do not, the rights holder may apply to a court for a limited injunction on the website.  

SOPA tries to incentivize the voluntary coming forward of ad and payment networks by extending them immunity from liability if they ”squeal” on an instance of copyright piracy or trademark infringement.

Finally, the bill expands the penalties related to certain activities including streaming video and various instances of counterfeiting.

Supporters of the bill tout it as the weapon American rights holders need to effectively police their IP rights. Supporting organizations include the obvious players such as the Recoding Industry Association of America (RIAA) and the Motion Picture Association of America (MPAA) but also large commercial entities like Nike, Ford and Pfizer.  SOPA would presumably make it easier for those large entities to enforce their trademark rights, adding to the ICANN Uniform Dispute Resolution Policy (UDRP).  While the latter only allows rights holders to go after infringing domain names, SOPA would allow them to pursue websites that make use of their trademarks in the actual content of the site itself.

Detractors contend that the bill is an absolute outrage and must be killed immediately. Representatives on both sides of the house have been outspoken critics, among them Rep. Nancy Pelosi and current Republican nominee for presidential candidacy Rep. Ron Paul.  In Rep. Paul’s own words, SOPA will likely result in "an explosion of innovation-killing lawsuits and litigation.”

Some even question the efficacy of the bill.  Sure the Attorney-General may order sites banned form the American web, but the DNS can be manipulated in creative ways to circumvent this roadblock. As was born witness to during this years “Arab Spring”, people can use mirror and proxy servers to get around national quarantines of the Internet.

The advent of these new measures also brings into question the utility of the DMCA’s notice and take-down system.  If SOPA passes, it will create a far stronger mechanism than what the DMCA currently offers thereby rendering those remedies obsolete. Unfortunately, the DMCA notice and take-down represents  significantly more balanced approach to dealing with piracy on the Internet than does the potentially Ex Parte (legal jargon meaning a hearing absent one of the parties) process put in place by SOPA. The latter allows actions to be taken against the infringing site itself if the Attorney-General is unable to locate the owner of the site.

It remains to be seen what type of effect SOPA and PROTECT IP will have in the Internet landcsape in the US.  The question also beckons as to whether any other countries will follow suit with national legislation of their own in the same vein.  Being the worlds largest exporter of Intellectual Property, the United States has a vested interest in seeing IP laws around the world tighten.  Here in Canada, our copyright modernization act (currently making its way through Parliament) seems to already reflect that exported interest.  The last thing Canada needs right now is “SOPA North”.  However, since “Stephen” and “Barack” are such good buddies these days, who knows what could happen?

Monday, December 5, 2011

Chanel gets TRO against 700 sites alleged to be selling counterfeited products infringing Chanel's Marks

A Federal Court judge sitting in the State of Nevada sided with Chanel when it asked to have approximately 700 domain names seized.  The sites allegedly contained counterfeit luxury goods including those of the plaintiff Chanel.  

Not only did the judge order the seizure of these domains, ha also ordered major social media sites like Google+ and Facebook and Twitter as well as search engines such as Google, Bing and Yahoo to de-index the targeted sites from their databases.

It seems that most of the investigatory work was done in house, though according to Ars-Technica, Brandon Tanori, a Nevada based private investigator, was called in to investigate a few of the sites in question.

Chanel is the owner of some 22 registered trademarks covierng a number of incarnations of is logo and company name. examples of some of the site names include:,, and

While the case has not yet gone to trial, Judge Kent Dawson granted the TRO (Temporary restraining order) requested by Chanel.  The TRO orders the domain name registrars currently in possession of the domain names in question to transfer those domains to an American registrar, who has been tasked with holding the domain names in trust for the court until conclusion of the pending action.

As mentioned above, the TRO also ordered search engines and social media sites to de-index the sites in question from their data-bases. The order makes no mention of the plaintiffs obligation to re-reimburse these 3rd party sites for the operational costs associated with the de-indexation.

While some have criticized the decision and evidence considered as being “one sided”, it should be noted that this judgement was rendered ex parte (meaning without all the parties present).  Normally, the rules of fundamental justice require that a party have the opportunity to appear and make it’s arguments known (the principle of Audi Alterem Partem).  The complaining party must at least serve notice of the action on the defendants.  

Given the nature of the case and the fact that the defendants are located around the world and that the information they provided to their respective registries may be false, the judge allowed for service by posting of a website:  Here, defendants can access PDF versions of all the court proceedings taken up to this point. This is a rather interesting new take on service by publication which many jurisdictions often use as an alternative when “personal” or “substitute” service are impossible. Service by publication usually requires leave of the court.

As pointed out by Ars Technica’s Nate Anderson (link above), decisions like this one may challenge the significance of the hotly debated Stop Online Piracy Act (SOPA). After all, the scope of the injunction in this case is incredibly broad.  

The judge is essentially commanding a host of large third parties to engage in a fair bit of work to enforce the court’s order. While the plaintiff was required to post a $20,000 bond, this sum is not slated to cover third party costs but represent a security on the damages that may be awarded the defendant should the court find them successful in arguing a wrongful injunction or restraint. If decisions like this one are to become the norm, one questions the need for strong legislation like SOPA.  

With respect, I tend to disagree in part with this view in so far as SOPA contains a number of other powerful measures such as restraining third party payment providers like PayPal or Moneybookers, and online advertising firms from supporting a given site or sites.  While the above mentioned point is well taken, it seems a little bit early to discount the eventual impact that SOPA and it’s sister the Protect IP Act will have on the legal landscape of the US, and by extentiton, the Internet as a whole.

Wednesday, November 23, 2011

A few talking points on bill C-11

 This is a version of a piece written for The Mark on some of the more noteworthy elements of bill C-11, The Copyright Modernization Act:

On June 2nd, 2010, the government introduced a new Copyright Modernization Act in the hopes of bringing Canada’s woefully out of date copyright law into the 21st century. Bill C-32, the third attempt in the last five years at achieving this, died on the order paper when parliament wad prorogued this past spring. Now, with a Conservative majority at the helm, the latest incarnation of Canada’s Copyright Modernization Act is speeding through the normal parliamentary process and is currently in its second reading. 

    The new Bill C-11 is more or less identical to Bill C-32. As such, it contains all the same features and pitfalls as the previous bill.  While Canada is desperately in need of a copyright makeover, a number of interest groups and individuals have questioned certain provisions contained in the bill and whether or not they will serve to benefit Canadians as a whole.
Photo by Renjith Krishnan

    The party line on C-11 (like C-32) is that the new amendments will provide Canada with a “balanced” copyright law that will adequately address the interests of all concerned parties while bringing an important element of Canadian law in line with the times.  Below are some of the main talking points of the bill that have stirred up controversy:

New Fair Dealing

In copyright law, there is a doctrine known as fair dealing which prescribes a list of categories where a person may freely make use of copyrighted content without the permission of the rights holder.  In other words, it allows people to carry on activities that, but for the fair dealing doctrine, would be clear cut instances of copyright infringement.

In current Canadian copyright law, research/private study, criticism/review and news reporting are the listed categories fair dealing. Bill C-11 seeks to add fair dealing for the purpose of education, parody and satire to the list.

Both supporters and detractors of the expansion of Canadian fair dealing have criticized the government’s approach.  Some supporters worry that the term “education” is too vague.  Intuitively, one might see this as a positive for supporters of educational fair dealing in that the lack of a strict definition may elicit a wider application.  While this may certainly be true, the reality of the situation is that mounting a fair dealing argument in defence of a copyright infringement law suit can prove to be a costly proposition. A more clearly defined statute may deter certain would be plaintiffs in situations where such clarity would have the effect of lessening their chances of success.

Detractors of educational fair dealing, most notably Access Copyright (the Canadian copyright licensing collective for reproduction of educational materials for all of Canada except Quebec which is under the auspices of Copibec) claim that such a provision will cause irreparable harm to authors whose materials are used in schools.

While parody and satire are not as hotly debated as educational fair dealing, there is one important question concerning these two new rights- namely, what’s the difference between them? The answer- which has proven to be less than obvious- changes depending on who is being asked. It seems that while parody denotes a more literal comical imitation, a satire is often described as more subtle.  Satire also often carries a social or political commentary where parody can be more “superficial”.  The fact remains that the difference between these two words is unclear and that they will most likely have to be defined by the courts.

Finally, proponents of the expansion of fair dealing argue that the list of categories should be non-exhaustive, allowing for new dealings to be considered fair as they arrive. This would mirror the approach in the United States. American “Fair Use” begins the list of categories with the words “such as” (see Title 17 Chapter 1 §107 USC) allowing for innovations in the technological and cultural marketplace to exercise a greater control over what is and is not “fair”.
New consumer exceptions

Believe it or not, it is currently an infringement of copyright law in Canada to rip a CD to your computer and transfer it onto your IPod.  It is also technically an infringement of copyright to use your “PVR” or “TEVO” to record shows for later viewing.  While these are both common practices in Canada, they have been, up until now, violations of copyright law.

Bill C-11 seeks to remedy that situation by allowing both “Format-Shifting” and “Time-Shifting” with the goal of legalizing these already common behaviours. While it is needless to say that cable and satellite providers such as Rogers and Bell have worked out contractual agreements insuring their customers with personal recording devices don’t get sued, it’s nice that the government thought to remove the need for such agreements by simply making these activities an exception to the scope of copyright.

User Generated Content

C-11 contains a novel provision never before seen in a national copyright law.  The Use Generated Content provision, a.k.a the “YouTube” exception, will allow non-commercial creators of “remixes” and “mash-ups” to make use of copyrighted content in the creation of their new derivative works.  This forward looking provision recognizes the paradigm shift in the creation and proliferation of culture described by Harvard Law Professor Lawrence Lessig as the move from “Read only” culture to “Read/Write” culture. 
The YouTube exception is, naturally, riddled with limitations and criteria. That being said, the simple fact that such a provision made it into the bill is a credit to Canada, a country often (rightly) criticized for having inadequate and out of date copyright law.

Anti-circumvention of “Digital Locks”

For all the reasoned and forward looking provisions mentioned above, C-11 contains another provision that has proven to be the chief object of criticism against the Conservative government in adopting the bill.

Clauses 41-41.22 of Bill C-11 which deal with the protection of “Technological Protection Measures” (TPMs) and “Digital Rights Management” (DRM) software disallows anyone from circumventing or otherwise “breaking” these protections. Briefly, a TPM or “digital lock” is essentially a piece of software that limits a user’s right to either access or copy a piece of media like a DVD or videogame. DRM software allows content distributers to manage these controls by monitoring the usage of the protected media and often transmitting this information back to the content distributer or a third party company.

These provisions have been included in the bill to make good on Canada’s treaty obligations pursuant to the WIPO Copyright Treaty (WCT) signed in 1996.  This treaty called for all signatory countries to include in their national copyright laws provisions that assure “adequate protections for digital locks”.  15 years later, C-11 seeks to ratify those treaty obligations.

Most of the criticism against the digital locks provisions is aimed at the fact that it trumps all the above mentioned rights, including fair dealing. While the WCT calls for “adequate” protection, some argue that the provisions contained in C-11 go far beyond the minimum requirements of the broad language used in the treaty.  Indeed many critics of the digital locks provisions, including University of Ottawa law Professor Michael Geist, have pointed out that even the American Digital Millennium Copyright Act (along with the accompanying case law developed since its inception more than a decade ago) is more permissive than C-11 on this point. 

The bill is being sped through Parliament and will likely be enacted exactly as drafted due to the majority government. Furthermore, the government conducted extensive consultation and examination during the committee process on Bill C-32 and is more than aware of all the interested groups’ positions on the subject.  One thing is certain; Canada’s copyright law is changing drastically. The next step will be to see how the courts interpret and apply this new law in the coming years.

Monday, August 15, 2011

CIRA Changes Domain Name Dispute Resolution Policy For .CA

For well over a decade now, Canadians have become accustomed to seeing “.ca” all over the web.  The Canadian web suffix is one of the fastest growing top level domains (TLDs) in the world.  For many domain registrants, .ca serves as a branding tool to designate a site or product as being “made in Canada”.  While the web may not have national boarders (at least not the type patrolled by customs agents), the argument can be made that a .ca designation serves as a key identifier and one Canadians (and those seeking Canadian content) use often to distinguish sites registered in Canada from the rest of the internet’s vast repertoire.

Photo by Svilen Milev
.ca is a “country code top-level domain” (ccTLD) and like any TLD must be overseen by a governing body. The Canadian Internet Registration Authority (CIRA) is the organization responsible for managing the .ca domain.  Established in 1998, CIRA took over the management of the .ca domain in 2000.  Until that time, it was overseen by a volunteer organization based out of UBC (lead by John Demco, former Computing Facilities Manager at the UBC Department of Computer Science).  According to CIRA, since it’s founding, over 1.6 million .ca domain names have been registered. 

Among the many tasks CIRA carries out, probably the most important is the implementation of its Domain Name Dispute Resolution Policy (CDRP).  This policy (an off-shoot of ICANN’s Uniform Domain Name Dispute Resolution Policy (UDRP)) is CIRA’s adjudication process by which trademark owners and businesses may enforce their marks against registrants who they believe are in violation of their rights. Businesses can also use this forum to combat cyber squatting- the practice of buying up domain names only to sell them to businesses who wish to register them at a higher price. 

The changes to the CDRP include:

“Rights” and “Use”: CIRA has seen fit to do away with these definitions in the policy.  On their web site CIRA explains that these definitions are more than anything a hindrance and create “overly technical and complex requirements in terms of what rights qualify for protection”.

Codification of “confusingly similar” analysis: CIRA has opted to standardize the test utilized by its panellists when determining whether a domain name is confusingly similar with a complainants mark. They’ve opted to use the “narrow resemblance” test over the traditional , broader confusion analysis of trademark law.  This test focuses on how much the domain name in question resembles the complainants mark in “appearance, sound and suggested idea” (see CIRA web site for more detail on the test).  

Changes to Bad Faith: If a complainant wishes their action to succeed, they must show, as per paragraph 4.1 of the CDRP, that the domain name was registered in bad faith (the act defines bad faith in paragraph 3.5).  The bad faith factors listed in the policy are now non-exhaustive- meaning that complainants may attempt to show bad faith on the part of the registrant in a manner not specifically contemplated by the policy.  CIRA has also added "use for commercial gain" to the list of factors mentioned in the CDRP.

Electronic filing: CIRA is now allowing parties to file all documents electronically in addition to hard copy filing.  If one wishes to file documents in hard copy, five copies must be made and faxed where one electronic copy is sufficient.  Wouldn’t it be nice if courts would follow suit...we can dream.

A more exhaustive list of modifications made to the CDRP can be found at the above link to CIRA’s web site.

Speculation on the changes has already crept into the blogosphere.  Professor Michael Geist, while praising some of the changes as being efficacious and appropriate to the Canadian context, had at least one criticism against the modification of the bad faith factors to become non-exhaustive:

“The exhaustive list was intended to guard against the ICANN experience where dispute panellists ventured well beyond clear cases of cybersquatting by creating their own categories of bad faith. Under the new CIRA policy, the bad faith list is now non-exhaustive, opening the door to more domain name dispute claims and increasing the risk of inconsistent decisions.”

Only time will tell if CIRA’s new moves will have the effect of streamlining the domain dispute resolution process.  The standardization of the confusion test should militate in favour of that reality while the opening of the bad faith factors may (according to Geist) hinder it. Whatever, the outcome, no one can rightly accuse CIRA of inaction.  

Sunday, July 31, 2011

Analysis of the Google Books settlement and its judicial rejection

The Google Books project and the suit

In 2004, Google undertook a gargantuan project to digitize a massive catalogue of books.  With the cooperation of a four university libraries and one public library, Google began scanning and cataloguing millions of books and volumes.  In the court’s written rejection of the Settlement agreement, the judge puts the number of books scanned at 12 million (though a recent estimate puts the number closer to 15 million).  

As part of the deal, Google would give the libraries a digital copy of their own catalogues and would create a master catalogue which would constitute the Corpus of the Google Books collection.  In typical Google fashion, this Corpus would be made searchable by the public.  For works in copyright, people would only be able to see snippets of the book containing the text of their search.  For works known to be in the public domain, the whole work would be viewable by the public.

Photo by Renjith Krishnan
To the objective observer this seems like an all around win for all parties involved.  Google grows its value while accomplishing a morally commendable enterprise; The public gains access to a wealth of culture and information otherwise lost or inaccessible; People with visual handicaps suddenly gain access to millions of works never before available to them; Authors of in-print books benefit from the muscle of Google’s impressive search algorithms to funnel otherwise untapped readers into purchasing their full work (once they’ve wet their appetites with the free snippet); Finally, authors of out-of-print books suddenly begin to see revenue again from material that was until now, commercially unavailable.  Indeed it seemed as though the Google Books project would bring the world closer to Shangri-La and “open knowledge”.  Certain people, however, saw things a little differently.

In September 2005, two lawsuits were brought against Google; One by the Authors Guild (Authors Guild v. Google) and the other by 5 major publishers (McGraw Hill v. Google).  The Authors Guild’s suit was a class action, ultimately joined by the publishers. 

The rights holders claimed “massive copyright infringement “on the part of Google.  Google contended that their acts were under the protection of the fair use doctrine.[1] Though public domain works were made available in their entirety, only snippets of protected works could be viewed for free. Google also specifically withheld the launch of ads on Google Books so that the claim couldn’t be made that they were engaging in commercial activity somewhat of a faux pas for a fair use defence.[2]

To this day, both parties stick by their original positions on the infringement/fair use issue even though such matters are no longer relevant (at least to this case). On October 28th, 2009, a settlement agreement between the plaintiff class and Google was reached.

The Settlement Agreement

The agreement has taken the form of a massive 166 page document (plus appendices) outlining the rights and obligations incumbent on Google and the plaintiffs.[3]  The agreement was originally supposed to cover all authors (including their heirs, successors and assignees) and publishers with an American copyright interest as of January 5th, 2009 (this scope was later changed.  See “International Law” below.).  

Google must also establish a “Book Rights Registry” where rights holders may register to receive royalty payments for their works that are included in the Corpus.  Google is to pay $34.5 million to establish and fund the operation of this registry.  

Not only that, Google must pay an additional $45 million into a settlement fund to pay out rights holders whose works were already digitized by Google without permission (as of May 5th, 2009).  In actual fact, the settlement mandates that each author who makes a claim be paid out a set amount making the $45 million a primary figure.  Any claims above that amount would still be paid out by Google as they arise and should the number be lower, the remainder is divided among rights holders, not returned to Google as surplus. 

In addition to these “sanctions”, the settlement gives Google extensive rights. The document lays out five:

  •  Google may continue growing the Corpus by digitizing books and inserts.   
  •  Google may sell subscription based access to the Corpus (such as institutional subscriptions for universities. 
  • Google may sell online access to individual books to users in an online store. 
  • Google may sell advertising on pages of books.
  • Other prescribed uses.
The settlement makes clear that those rights are non-exclusive and can therefore be licensed by rights holders to anyone, including direct competitors of Google. 

Google must share the revenues from these uses giving up 63% to the rights holders for works published before January 5th 2009. For new works going forward, Google must hand over 70% of revenues from all sales (both subscription and per-use based) and advertising revenue derived from the Corpus- less a 10% deduction to cover Google’s operating costs.

The settlement draws an important distinction between two classes of books included in the Corpus: In-print books and out-of-print books.  In-print books are those that are still commercially available.  Concordantly, out-of-print books are those that are no longer being produced and are not commercially available.  An interesting question arises when a publisher offers online versions of books that it no longer prints for retail.  Is that book still in-print? It is commercially available but only in a digital format. 

It’s also worth mentioning that the settlement agreement stipulates the amount to be paid to the lawyers representing the class is $45.5 million.  Though this consideration is immaterial to the meat of the issue, it’s still a tad disconcerting that the lawyers are making more than the amount being paid out to establish a registry for the authors of what some say is up to 15 million books and counting.

Rejection of the Settlement by the Court of the Southern District of New York

On March 22nd, 2011, Judge Chin of the Court of the Southern District of New York rejected the settlement agreement that received preliminary approval five months earlier.[4]  Judge Chin disagreed with Judge John E. Sprizzo on the fairness of the settlement agreement.  In truth, Judge Chin had the benefit of reading the hundreds of objections that were filed with the court between the preliminary ruling and his ruling. 

Before stating his reasons for rejecting the settlement, Judge Chin admitted two factors considered that weighed in favour of approval.  First, he commended the “arm’s length negotiation between experienced, capable counsel, with assistance from the DOJ” (Department Of Justice). Second, he rightly asserted that proceeding with a full blown trial (and the inevitable appeals process) would be very lengthy and wildly expensive.

That being said, Judge Chin had a number of valid reasons for rejecting the proposed settlement:

·         Inadequate representation of the class members: The class of plaintiffs in the case is very large.  Google sent 1.26 million notices in 36 languages to copyright holders and potential class members as well as publishers and authors’ rights collectives.  For example, Scholars and academics usually publish with a motivation that is very different from a commercial writer.  Both are included in the settlement.  It is even conceivably arguable that publishers have very different motivations from authors and authors’ rights collectives. 

·         Encroachment by the court on Congress’ power: Under the settlement agreement, Google would have the rights to all orphan works[5] without any prior consent (a rights holder may step forward after the fact and make a claim).  Judge Chin acknowledged Sony Corp. Of America v. Universal City Studios Inc[6]. where the Supreme Court held that it is congress’ duty to keep up with technological advances and their effects on copyright law.  He further cited Eldred v. Ashcroft[7] where the Supreme Court said “It is generally for Congress, not the courts, to decide how best to pursue the Copyright Clause’s objectives.”

·         Google should not be allowed to take a “shortcut”: Instead of undergoing the long and prohibitively costly process of licensing and tracking down of rights holders, Google decided to take a “sue me” shortcut.  In violating copyright on an epic scale, they hope to circumvent both the time and cost associated with the above scenario.  Ironic that seven years and millions of dollars in billable hours later, we still have neither a verdict, nor a settlement.  As it was eloquently put by Stanford Law School Professor Pamela Samuelson “We’re giving Google a blank check to essentially engage in activity that would be considered clearly infringing activity but for the settlement”

·         Anti-trust concerns:  The settlement agreement would have the effect of giving Google the exclusive monopoly right over all orphan works.  This means that any institution or competitor who wishes to make use of or offer any of the millions of orphan works in the corpus cannot do so without paying Google a royalty.  Google also has the right to refuse a license to anyone it sees fit.  This is a prime example of what anti-trust law is supposed to prevent- anti-competitive and monopolistic behaviour.  The fact that this is a settlement for a legal action adds insult to injury in that it would create an injustice, potentially greater than the initial reason Google was sued in the first place.  There’s no Fair Use defence in anti-trust for Google to hide behind either.

·         Privacy concerns: No one is better at collecting data than Google.  Would this extend to information about the books we’re reading? Not only will Google know what you’re reading and when, they’ll know for how long, how many pages and which pages you read.  This is clearly a breach of our reasonable expectation of privacy.  What we read can often be personal.  It isn’t the sort of data everyone is ok with others collecting.  There is also the fear that such information can be forcibly turned over to government entities by way of the aging Electronic Communications Privacy Act (ECPA).  If Google Books ever does get started, don’t be surprised to hear about the F.B.I or the Department of Homeland Security compelling Google to hand over reader data of certain subscribers.[8]  Never before has law enforcement been so well equipped with readily available and invasive access into our lives (e-mails, online profiles and even cell phone location data).  Google Books subscriber data would simply be one more tool available to them.

·         International law: At first, the settlement was worded to include any book that had a U.S. copyright interest.  This would include almost every single book in the world.  The U.S is a signatory to the Berne Convention.  This means that the U.S. must give equal recognition and protection to any work published or performed in the U.S. regardless of its origin.  Objections to this language being used in the settlement resulted in the definition being narrowed to exclude non-American works that were not affirmatively registered in the U.S. Copyright Office.  That being said, Canadian, British and Australian works were still included in the settlement regardless of U.S. copyright registration. 

Uncertainty ahead

With this rejection on the books, it’s back to the drawing board for Google and the plaintiffs.  Though impressive opposition to the settlement has been mounted, it is unclear whether it will be abandoned entirely or not.  

The judge alluded to the fact that a matter like this is one for Congress, not the court to decide.  Which beckons the question, can Congress really do any better? There are mixed opinions as to how that question should be answered. However, one might posit that regardless of one’s faith in Congress to come to a better result, the Constitution mandates it.  

Article 1 Section 8 Clause 8 of the U.S. Constitution gives Congress the power to “To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.”[9]

The Settlement has far reaching implications on an entire category of works contemplated in the Copyright Act, orphan works.  This fact alone should militate in favour of a Congressional rather than judicial response.
On its face, this settlement appears to be a misuse of the class action mechanism.  In general, a class action suit is usually taken to compensate the victims of damages caused by the plaintiff(s).  Though the settlement covers compensatory considerations, it also provides for a forward looking licensing and royalty scheme. This scheme also implicates millions of people who some contend are not being accurately represented by the class representatives. 

In a debate on the Settlement at the University of Richmond Law School, Professor James Grimmelmann of New York Law School delivered a brilliant analogy to illustrate this point:

Say that 5 years ago there was a minor oil spill (nothing on the magnitude to the one that recently occurred) in the Gulf of Mexico caused by BP.  A class forms that comes to represent all residents and business of Gulf Coast states.  In a settlement, to facilitate the pay-out of any future damages caused by an oil spill in the area, BP sets up a compensation fund and a streamlined process by which people can easily make claims.  Imagine that BP agrees to put $500 million in that fund to assure that no damage that could possibly arise would remain uncompensated.  Fast forward to the events of last year.  That settlement, already having been signed and approved would exonerate BP from liability for the astronomical damages it has caused to date (far exceeding the once seemingly adequate sum of $500 million).[10]

This scenario helps highlight exactly what may be wrong with a significantly forward looking settlement like this one.

Professor Samuelson brings up an interesting point on the future implications of the settlement on the Corpus itself.  What if the settlement goes through and 10, 15 or 20 years down the line Google shuts down or goes bankrupt or otherwise?  We will all have become so dependent on Google Books that its failure would become an unacceptable reality.  A service like this truly is a public good and as such, should be protected.  This presents a “too big to fail” scenario in which the Government may eventually have to step in out of necessity.  

Google also has the right to sell the Corpus to anyone.   Some people fear that Google has carte blanche to engage in “price-gouging” considering they are the only source for a large part of the Corpus. Even if we presume Google will not abuse their right, who’s to say the next owner wont? As Samuelson suggests, those who aren’t afraid of price-gouging on the part of Google may reassess that fear should the database be sold in the future. 

This entire saga started out with Google trying to do what Google does best, cataloguing and indexing data in order to make it searchable.  The matter has evolved into what could be one of the most important and impacting copyright cases in history.  While experts continue to speculate on possible and desirable outcomes, people concerned with copyright around the world are watching intently as the situation continues to slowly unfold.  

[1] Fair use is a part of the American Copyright act that presents a non-exhaustive list of uses of a work that would otherwise infringe copyright but do not because they are considered to be fair uses of the original work.  Canadian Copyright law has a concurrent (but not identical) regime referred to as fair dealing.  These regimes are not to be viewed as a simple defence, but as an integral part of the copyright act and therefore a clear delimitation on the exclusive right afforded a copyright holder.  See McLachlin CJ. In the Supreme Court of Canada’s decision in CCH Canadian Ltd. v. Law society of Upper Canada: “...the fair dealing exception is perhaps more properly understood as an integral part of the Copyright Act than simply a defence.”
[2] One of the factors considered in the Fair Use analysis is the purpose and character of the use.  It is harder to mount a fair use defence if the use in question is highly profitable to the alleged infringer.  That being said, the commercial nature of a use does not automatically preclude it from being fair.  The U.S. Supreme Court has even said that this factor is not the most important to be considered.
[3] Google makes the updated settlement agreement as well as an FAQ on the settlement available at the following link: <>.
[5] An Orphan Work is one where the Copyright owner is unknown or cannot be located but is still under copyright protection.  One problem relating to Orphan works is that if they are used without clearing the rights, a task that cannot be completed, the user opens themselves to the risk of the rights holder “popping up” out of nowhere with an infringement suit.
[6] 464 U.S. 417 (1984). For the full text of the decision see: <>
[8] For more information on the ECPA and it’s implications on online privacy, please see: <>

Sunday, July 24, 2011

Copyright in Databases?

Can one have copyright in a database? For those familiar with the law, it is clear that raw data itself is not protected by copyright. If it were, the phonebook (or whoever else was “first to fix” my information) would have a copyright on my name, address and telephone number.  

Photo by Renjith Krishnan
Databases are considered “compilations” under copyright law.  Both the Canadian and American Copyright Acts grant copyright protection to the compiler. Section 2 of the Canadian act defines a compilation as: “a work resulting from the selection or arrangement of literary, dramatic, musical or artistic works or of parts thereof, or a work resulting from the selection or arrangement of data” (Emphasis added).  This means that the data collected in the database need not be original or creative.  A list of names and numbers may be eligible for protection as a compilation.  The copyrightable element of the work is the “selection and arrangement” on the part of the compiler. 

For example, last year, Professor Michael Geist put together a collection of scholarly, peer-reviewed articles on Canadian copyright and the (then) new bill C-32 (electronic version available free at the above link).  Each chapter comprised a paper written by a different academic.  Some of the articles contain a very short forward by Geist.  In this case, professor Geist has copyright in the compilation.  Though he didn’t write the majority of the collection, he was the organizer (or to use the terms of the law the “selector and arranger”) of the content.

How much creativity or effort a person must put into this selection process is not readily discernable.   In the 1991 American Supreme Court decision known commonly as Feist, the principles of which have been adopted in Canada, the court solidified the applicable test in an effort to demystify this standard.

The court’s decision in Feist comes down to a two part test:
  • Is the material (selection and arrangement) copyrightable in favour of the compiler?
  • Has the defendant reproduced an infringing amount of the copyrightable material?

The answer to the first question varies from one jurisdiction to another.  There are essentially two competing doctrines to determine the copyright-ability of a work: the “creativity” approach and the “sweat of the brow” approach.  

In jurisdictions like Australia and the UK, a compiler will be granted copyright in their database if they dedicated a substantial amount of time, money and/or resources to its completion. This has largely been the Commonwealth tradition of which Canada is historically a part. 

The creativity approach demands that the person claiming copyright in their database demonstrate that they exercised a minimal amount of creative energy in selecting and arranging the data.  This is the approach held up in Feist which had to do with a telephone directory.

In that case, the appellant Feist was a company whose business it was to compile large telephone directories from smaller local directories.  The plaintiff, Rural, sued Feist for copyright infringement and lost due to there having been no originality whatsoever exercised by them in the selection and arrangement of their data.  The court tells us that originality (by way of creativity) has always been the cornerstone of copyright in the US. That being said, the threshold for creativity is very low.  One must simply demonstrate a spark of originality.  For example, alphabetically listing the names and phone numbers of everyone living in a geographical area is totally unoriginal. However, perhaps a register of restaurants in a designated geographical area listed by style or genre would be eligible.  To pass the creativity test, it is unnecessary that the compiler show true novelty or non-obviousness as in patent law.  They must simply show that “spark” the court speaks of so romantically.

As mentioned above, Canada has historically followed its commonwealth roots and the “sweat of the brow” approach. However, the precedent setting Canadian case in this area shows great consistency with the American approach.

In Tele-Direct v. ABI, the Federal Court of Appeals upheld the Trial Court’s decision that the elements of Tele-Direct’s Yellow Pages directory were not in themselves sufficiently original.  While ABI admitted from the outset that Tele-Direct had copyright in the overall compilation, the court found that insufficient originality was exercised in the selection and arrangement of the “sub-categories” (the information in each individual listing).  ABI didn’t copy the overall form of Tele-Direct’s Yellow Pages; they simply copied the data in the individual listings, which were themselves, not organized or arranged in a sufficiently original manner to be worthy of copyright protection.

This break with the Commonwealth trend is probably due in part to Canada’s adherence to NAFTA (North American Free Trade Agreement) and TRIPs (trade-related aspects of intellectual property rights). In his 1998 submission to Industry Canada and Canadian Heritage, Professor Robert Howell highlights article 1705(1)(b) of NAFTA which reads as follows:

“1. Each Party shall protect the works covered by Article 2 of the Berne
Convention, including any other works that embody original expression within the
meaning of that Convention. In particular:
. . .
“(b) compilation of data or other material, whether in machine readable or other
form, which by reason of the selection or arrangement of their contents constitute
intellectual creations, shall be protected as such” (emphasis added).

He further purports that the intention in defining compilation this way was likely to coincide with the definition given in the 1976 US Copyright Act: “A ‘compilation’ is a work formed by the collection and assembly of pre-existing materials or of data that are selected, coordinated, or arranged in such a way that the resulting work as a whole constitutes an original work of authorship. The term ‘compilation’ includes collective works.” (emphasis added).

In the EU, databases benefit from the protection of two legislative regimes.  As in North America, database creators in Europe may avail themselves to regular copyright protection. They may also make use of the “European Database Directive” (DIRECTIVE 96/9/EC). This Sui Generis regime doesn’t require the database creator to have invested any level of creativity or originality in his work.

Article 7 section 1 of the directive reads as follows:

“Member States shall provide for a right for the maker of a database which shows that there has been 
qualitatively and/or quantitatively a substantial investment in either the obtaining, verification or presentation of the contents to prevent extraction and/or re-utilization of the whole or of a substantial part, evaluated qualitatively and/or quantitatively, of the contents of that database.” (emphasis added).
This regime clearly endorses the “sweat of the brow” approach. Though some may disagree with this standard, the Directive comprises another provision that is by far more controversial and altogether unsettling.

Though Article 7 section 1 of the directive says that the term of protection afforded the database under the law is 15 years, section 3 of the same article reads as follows:

“Any substantial change, evaluated qualitatively or quantitatively, to the contents of a database, including any substantial change resulting from the accumulation of successive additions, deletions or alterations, which would result in the database being considered to be a substantial new investment, evaluated qualitatively or quantitatively, shall qualify the database resulting from that investment for its own term of protection.”

This essentially means that by making simple alterations, additions or subtractions of content or structure, the term resets for another 15 years.  This amounts to what is clearly “evergreen” protection for databases in the EU.

While this directive has been going strong in Europe for the past 15 years, neither the US or Canada have enacted similar legislation, most likely because they aren’t convinced it’s necessary to do so.

As the EU continues to distinguish itself in the area of database protection, the west seems to be coming together.  Though originality with reference to copyrightable works is constitutionally mandated in the US (Article 1, section 1, clause 8), no such Constitutional obligation exists in Canada.  That being said, both international agreements and the interest of uniformity have seen Canada and the US align closely on the judicial treatment of copyright in databases.  It is my belief that with the monumental growth of cloud computing and online data storage, congruity in this area may prove beneficial to the attraction of American companies who wish to be certain that they may obtain Canadian copyright in their property.