Shifting formats - the IPO’s draft private copying exception

[Originally published 16 July 2013, based on the draft circulated for technical review.  Updated 14 April 2014 to take account of the amendments contained in the draft Personal Copies for Private Use Regulations laid before Parliament on 27 March 2014.]

The draft private copying exception (PDF) released for consultation on 7 June 2013 is one of the UK Intellectual Property Office’s flagship copyright reforms following the Hargreaves Review.  In his final report (PDF) in May 2011 Professor Ian Hargreaves focused on the lack of a format-shifting exception as particularly threatening to the legitimacy of copyright.  He identified:

“a growing mismatch between what is allowed under copyright exceptions, and the reasonable expectations and behaviour of most people. Digital technology has enabled use and reuse of material by private individuals in ways that they do not feel are wrong – such as sharing music tracks with immediate family members, or transferring a track from a CD to play in the car. It is difficult for anyone to understand why it is legal to lend a friend a book, but not a digital music file. The picture is confused by the way some online content is now sold with permissions to format shift (iTunes tracks) or to “lend” files (Amazon ebooks) at no extra cost.  This puts the law into confusion and disrepute. It is not a tenable state of affairs.”

If respect for copyright is to be restored, the proposed new private copying exception has to shoulder much of the work. 

It has to succeed on several fronts.  The audience for many aspects of copyright is copyright professionals.  However, since the private copying exception will determine what everyday users of PCs, laptops, tablets and mobile devices can and cannot do with music and movies that they purchase, its audience is the general public.

As such the exception needs to be clear, certain and sensible.  For clarity, it should not require special expertise in copyright law to understand the exception and its practical implications.  For certainty, there should be as few as possible areas of doubt about what is and is not permitted.  As to being sensible, there should be a minimum of arbitrary distinctions.  Not only should the user be able readily to understand which particular activities fall on one or other side of the line, but the reasons why the line is drawn where it is should be coherent and readily apparent.  In general terms the boundaries should accord with the reasonable expectations of users, albeit it must be conceded that there are legitimate areas of disagreement as to where the boundaries should lie.

If the exception does not pass these tests, not only will an opportunity to set copyright back on the road to legitimacy have been lost, but further damage to copyright will be done as another specimen of obscurity, arbitrariness and remoteness from reality is added to the list.  In an age when policymakers demand that users should be exhorted to respect copyright, it behoves our legislators to write laws that those users can readily understand and which, when they have understood them, make sense.

This article evaluates the published draft of the exception [now the draft Statutory Instrument laid before Parliament] with those criteria in mind.  We invoke a hypothetical user, Bruce, and try to decide which of his everyday format-shifting activities would or would not be permitted by the proposed new S28B of the 1988 Copyright Designs and Patents Act. 

For the most part the article avoids hotly contested issues about how broad or narrow the exception should be, whether its implementation without a corresponding copyright levy is or is not compliant with the EU Copyright Directive, and whether the whole thing is undermined anyway by the prohibition on circumvention of technical measures. 

The reader can make up his or her own mind whether the draft exception passes the tests of clarity, certainty and sensible boundaries.  To the extent that it may not, we should spare a modicum of sympathy for the Intellectual Property Office since some uncertainty is the inevitable result of lack of certainty in the EU Copyright Directive, which constrains the UK government's freedom of action. 

What follows is the tentative result of wrestling with the short but not entirely simple text of 28B.  Others may well have different views on how 28B is meant to apply.  It would certainly be welcome to discover that I have overcomplicated the thing and that it really doesn’t take the 4,000 words of this piece to analyse it.  But I am not optimistic.

In any event at present 28B is draft legislation, open to consultation until 17 July 2013.  The draft will quite likely be amended before anything is put before Parliament.  [Indeed the SI laid before Parliament is quite significantly amended.]

Preliminary

28B divides logically into two steps.  The first step is to identify whether Bruce has suitable source material at hand.  28B defines this as a copy of a copyright work lawfully acquired by Bruce and which he holds on a permanent basis (for example it is not a copy that is rented to the individual for a specified period or borrowed from a library).  [Eligible source material is now to be called an "individual's own copy". That is a copy which Bruce has lawfully acquired on a permanent basis, is not an infringing copy and has not itself been made under a copyright exception. Lawfully acquired on a permanent basis specifically includes a purchase, a gift, or a download resulting from a purchase or a gift (other than a download enabling no more than temporary access to the copy); but does not include a borrowed, rented, broadcast or streamed copy. None of this will now apply to computer programs, which are excluded entirely from 28B.]

The second step is to consider what use Bruce can make of that source material.  In general terms he can use the source material to make a further cop[ies]y of the work provided that the further cop[ies]y [are]is made for Bruce’s private use for ends that are neither directly nor indirectly commercial. Bruce cannot transfer the further cop[ies]y to another person[, other than on a private and temporary basis]; and if he transfers the source material to another person [(other than on a private and temporary basis)] without destroying the further cop[ies]y, that is an infringement. [A further copy made under 28B is now to be called a "personal copy"]

That may sound fairly simple.  However subtle complications arise when we try to apply it to the multiplicity of real world scenarios, including some specifically mentioned by the Hargreaves Report.

For the sake of relative simplicity we have not addressed the effect (if any) of contractual prohibitions, and have ignored the effect of any TPM restriction. [28B will render a contract term unenforceable to the extent that it purports to prevent or restrict the making of a copy which would under 28B would not infringe copyright. TPMs will remain effective.] 

Source material

First, consider the source material that Bruce must have in order to be able to use 28B.  Would source material that Bruce acquired in any of the following ways qualify?

[Remember that none of this applies to computer programs and the source material cannot itself have been made under an exception from copyright.]

(a)  Bruce purchased a genuine CD

Definitely OK.  28B can be used where the individual has ‘acquired’ a copy lawfully on a permanent basis.  Purchasing a genuine CD is the paradigm case.  28B could also be used if Bruce had received the genuine CD as a gift, either new or second-hand, or had bought it second hand. But Bruce could not use 28B if he had rented the CD or borrowed it from a library, from his sister or from a friend. [But not computer programs. Computer programs have their own regime of copyright exceptions under the EU Software Directive. See sections 50A to 50C of the 1988 Copyright Act.]  

(b)  Bruce purchased an infringing CD

Copyright is infringed by the person who made the infringing CD; and also by the person who sold it if he knew or has reason to believe it was infringing.  Bruce did not infringe copyright by merely purchasing it.  So from Bruce’s perspective as a matter of copyright law he may have acquired it lawfully, even though it was an infringing copy.  This might suggest that he can use it as source material for making a 28B copy.

Contrary to that, in its December 2012 response to consultation the government said that the exception would be limited to use by people who already own a lawful copy of a work; and that the exception would permit copying only by a lawful owner of an original copy.

But differently again, in its recommendation in the same document the government said the exception would allow a lawful owner or buyer of a copy of a work to reproduce that copy for their personal use. 

Lawfully acquiring a copy and lawfully acquiring a lawful copy differ significantly from each other.  It is not obvious from the consultation history which the government intends. [Now clarified.  28B now makes clear that the source material cannot be an infringing copy.] 

As to the merits of each approach, on the one hand why should someone be able to make use of 28B to make another copy of something which already infringes?  On the other hand, why should the consumer be at risk of infringement simply because the original copy turns out to be infringing?  The consumer is poorly placed to know with any certainty whether what s/he has purchased is genuine or infringing.  That is especially the case with secondhand CDs, but can also apply to apparently genuine original CDs.  (It can happen that a well known label sold through reputable retailers turns out to be unlicensed.)  Would it be fair to the consumer that his or her ability to use the exception depends on something that will often be out of her knowledge and control? Should the exception be inapplicable only where it was obvious to the consumer that s/he was acquiring an infringing item? [It will make no difference if Bruce has no means of knowing that the source material is an infringing copy.]

It is also unclear whether private copying from an infringing source copy is permitted under the EU Copyright Directive, with which this exception must comply.  Whether the private copy exception under the Directive can apply to unlawful source copies is the subject of a currently pending reference to the EU Court of Justice, in ACI Adam and Others Case C-435/12.  [Decided 10 April 2014.]

(c)  Bruce borrowed a copied CD from a friend

No.  A borrowed CD, whether copied or genuine, is not held permanently.  The exception could be used to make a private use copy only from a copy held on a permanent basis.  [No change.]

(d)  Bruce received a copied CD as a gift

If the copy for some reason does not infringe, then OK.  Bruce holds it on a permanent basis.  If it does infringe, then the position is the same as scenario (b): unclear. [Now clarified. As in (b), an infringing copy cannot be used as source material for 28B.]

(e)  Bruce purchased a permanent legitimate copy of a song by download over the internet. 

OK, or at least so the government appears to intend.   It mentioned eBooks in the December 2012 impact assessment.  In the Dec 2012 consultation paper it said “This would include downloaded digital content that has been purchased to keep (or licensed in an analogous way) such as eBooks” (Consultation note 41).

However the actual wording of the exception may not be quite so clear.  Unlike acquiring a CD, downloading involves making a new copy.  Does ‘acquiring’ include making a new copy? Since the government apparently intends that it should, it would be as well to put this beyond doubt.

In any event the exception does not apply to rented or streamed materials. [Now clarified.  Bruce will be able to use a copy acquired by means of a non-infringing download resulting from a purchase or a gift (other than a download enabling no more than temporary access to the copy). Bruce will not be able to use borrowed, rented, broadcast or streamed source materials.]

Use scenarios

Let us assume that Bruce is equipped with a permanent copy that he can use as source material under S28B.  We will assume that it is a genuine purchased CD (not unlawfully copied, not an unlawful parallel import from outside the EEA).  Would any of the following acts be permitted under S28B?

(a)  Bruce copies the CD to the hard drive of his PC. 

OK, so long as he makes the copy for his private use for ends that are neither directly nor indirectly commercial.  However, whether he has made the copy for his private use turns out to be quite a tricky question.  Some uses involve further copies being generated.  Others do not (or at least only transient copies).  How does 28B apply to each situation? 

28B legitimises only a copy made for Bruce’s own private use.  So the key question appears to be what was Bruce’s purpose when he made the copy?  That appears to depend at least in part on what was in his head at the time he made the copy (there may be room for arguments about whether the test is subjective, objective or hybrid – we will assume that Bruce’s intention has some relevance).  However surrounding circumstances would offer some pretty big clues to the purpose.  [28B now fleshes out the concept of private use with some examples. It includes private use facilitated by:

- making a copy as a back-up copy
- making a copy for the purposes of format-shifting
- making a copy for the purposes of storage (including in an electronic storage area accessed by means of the internet or similar means which is accessible only by the individual (and the person responsible for the storage area).]

Here are some sub-scenarios:

(i)   Bruce copied the CD into a folder accessible only to himself.

OK. That would be consistent with Bruce’s professed private use purpose.

(ii)  Bruce copied the CD into a public P2P sharing folder on his hard drive.

Probably not OK.  Since it is accessible by members of the public Bruce would have an uphill task convincing anyone that the copy was made for his private use. 

(iii)  Bruce copied the CD into a folder accessible to a small circle of others (such as family members) on his home network. 

If Bruce did this with so as to enable others on his home network to make their own permanent copies, that is beyond Bruce’s own private use and would not be legitimised by 28B.  [No change. Although Bruce will be able to 'daisy-chain' (see below) only he will be able to make further personal copies.]

If Bruce did it so as to enable others on his home network independently to play and listen to the tracks in the folder, that is probably not legitimised by 28B, since their playing and listening to the tracks would be private use by them, but not by Bruce.

If Bruce did it with the intention only of listening to the tracks himself, that might be OK since that is Bruce's own private use.  However Bruce could be challenged as to why, if that was his intention, he put the tracks in a folder accessible to others.

Establishing Bruce’s true intention could be very difficult.  It might depend at least partly on his technical sophistication, such as whether he knew that others could access the folder to which he copied the CD.  That might be especially the case if some other member of the family set up the network and configured his computer. 

(iv)  Bruce copied the CD into a folder accessible only to himself, but with the intention of playing the tracks from the folder through his home loudspeakers via a streaming device.

Unclear.  If Bruce intended to listen through personal headphones that would clearly be his private use.  Playing the tracks through his home loudspeakers means that other domestic and family members can listen to the music that Bruce is playing.  Is this also Bruce’s private use? 

The concept of private use as understood in EU law is wide enough to cover domestic and family use.  Is that what 28B means?  If so does the fact that the playing is under Bruce’s control mean that it counts as Bruce’s private use, as required by 28B?  28B itself offers no illumination, other than that using the word 'private' rather than 'personal' use may suggest that this is OK. (Confusingly however, the IPO uses the term 'personal use' in its commentary on the draft exception).

If it is not OK, then the reasonable user could well regard the results as arbitrary.  It would be permissible to buy a CD and play it to friends and family on a music system at home; but not permissible to copy the CD to a computer or an MP3 player and play that copy to the same audience through the same music system.  [No change. None of the supporting documents or the Impact Assessment shed definitive light on this scenario, although the tenor of the rejected Impact Assessment Option 2 may suggest that the answer is 'Not OK'.]

(b)  Bruce copies the copy that he made in scenario (a) from his hard drive to his portable device (e.g. MP3 player, mobile phone or USB stick)

This 'daisy-chaining' is a very common scenario.  The Gowers Report back in 2006 recognised the importance of daisy-chaining for format shifting: "The exception would only allow one copy per ‘format’, but it would also have to recognise that transfer between formats may require intermediate steps (or formats) to be taken." (para 4.76)

If the copy made from the CD to the hard drive in scenario (a) was legitimised by 28B, this further copy made from it is probably also legitimate if made for Bruce’s private use.  But 28B does not spell this out beyond question. 

Since we are assuming that Bruce legitimately made the copy on his hard drive under 28B, he has lawfully acquired that copy (if we are right that lawfully acquiring includes lawfully making a new copy).  Under 28B he is entitled to keep the hard drive copy permanently unless and until he transfers the original CD to another person.

On that basis Bruce can probably ‘daisy chain’ 28B to make successive copies from the CD to his hard drive and then from his hard drive to his portable device, so long as both copies are for his own private use.  He could presumably keep both copies and continue to sync between them.

Assuming this to be the government's intention, it would be better if 28B made completely clear that daisy chaining is permitted.  [Clarified.  Bruce can make a personal copy under 28B from an existing personal copy that Bruce (but no-one else) has already made under 28B.] 

If Bruce intends to use the portable device to play the tracks in the presence of others in his private sphere, such as in his car (the example given by Hargreaves), then the same uncertainties about private use arise as with the hard drive copy. [No change.]

(c)   Bruce copies the copy that he made in scenario (a) from his hard drive to his private cloud storage area

OK, if daisy-chaining is permissible.  28B places no restriction on the place in which the 28B copy is stored – only that it be made for his private use for ends that are neither directly nor indirectly commercial.  On the face of it that would include a private cloud account.  Since storing the copy in that account involves making a new copy, the ability to daisy-chain is important in order to bring the cloud copy within 28B.  [Daisy-chaining will be OK. See above.]

Cloud storage is specifically addressed in 28B(3).  The drafters appear to have been concerned that storing in the cloud could amount to a transfer of a copy to another person, and thus trigger the infringement provisions of 28B(2).  They have therefore provided that nothing in 28B(2) prevents an individual from storing a further copy made under 28B in an electronic storage facility accessed by means of the internet or similar means, where that facility is provided for his sole private use. 

While specific provision for cloud storage is probably necessary from the point of view of clarity, this sub-section is problematic.   First ‘storing a further copy made pursuant to [28B]’ is an odd phrase in his context, since the act of storing a copy in the cloud necessarily entails the making of an additional copy. Is the section referring to that further copy, to a pre-existing further copy (e.g. made by copying from CD to computer hard drive) or to both? 

The uncertainty is compounded by the reference to 28B(2), since that deals with transferring “the copy”.  That most obviously means the transfer of a pre-existing copy – such as when Bruce gives a USB stick containing a copy to his sister.  On that understanding storage in the cloud would not be a transfer of ‘the copy’ at all, since it involves the making of a new copy in the cloud.  Yet 28B(3) seems to suggest that this process could be understood as a transfer of "the copy". 

If “transfers the copy” is indeed broad enough to include a transfer process that creates an additional copy, that introduces further complications into the operation of 28B(2), since more copies will come into existence than 28B(2) appears to contemplate. [Re-drafted and clarified.  Bruce's copy on his hard drive made from his CD is a personal copy under 28B.  He can make a further personal copy from that and store it in his private cloud storage.  However this only applies if the storage area is accessible only by Bruce and the person responsible for the storage area. Shared storage areas will not count.]

(d)  Bruce copies the copy that he made in scenario (a) from his hard drive to his sister’s portable device for her.

No.  The second act of copying (from hard drive to his sister’s device) would not be for his own private use.  [No change.]

(e)  Bruce copies to his own MP3 player with the intention of lending the MP3 player to his sister.

No, at least if Bruce’s purpose in doing this was that his sister should be able to listen to the tracks.  The copy was not made for Bruce’s private use, but that of his sister.  The same would apply if Bruce copied to a USB stick with the intention of lending it to her.  [No change.]

(f)    Bruce, having copied to his own MP3 player for his own private use, later decides to lend the MP3 player to his sister.

Probably OK?  Since the later decision to lend the MP3 player was not Bruce’s original purpose, that does not prevent the copy being originally legitimate under 28B.  There is nothing in 28B to suggest that it becomes illegitimate if Bruce's intention changes later.  As Bruce is only loaning the MP3 player, it is not caught by the prohibition on transferring (otherwise than on a private and temporary basis) the 28B copy to another person.  The same would apply to a USB stick. [No change.] 

It would be a separate question whether, when the sister listens to the tracks stored on the borrowed device, she infringes copyright.  Probably not (so long as she does so in private circles), but that may be affected by the outcome of a currently pending reference to the CJEU in the Meltwater litigation. 

(g)  Bruce, having copied a CD to a USB stick for his private use, then gives the USB stick to his sister.

No.  Although the USB stick copy was made legitimately under 28B, under 28B2(a) Bruce infringes copyright if he transfers that copy to another person otherwise than on a private and temporary basis.  That previously legitimate 28B copy becomes an infringing copy.

(h)  Bruce, without making any copy of it, gives a genuine purchased CD to his sister.

OK.  Under general copyright law Bruce is entitled to transfer to someone else a physical copy that was previously put into circulation in the EEA by or with the consent of the rightsholder (the principle of ‘exhaustion of rights’). [No change.]

(i)    Bruce, having copied a genuine purchased CD onto a USB stick for his brother, then gives the CD to his sister.

Copying the CD is outside 28B as it is for Bruce’s brother.  That copy therefore infringes.  Bruce can give the genuine CD to his sister since exhaustion of rights still applies to the CD. [No change.]

(j)    Bruce, having copied a genuine purchased CD onto a USB stick for his private use, then gives the CD to his sister.

OK, but only if Bruce first destroys the copy on the USB stick.  If not, he infringes copyright by transferring the CD to another person.  However it is the copy on the USB stick, not the CD, which is then regarded as an infringing copy.  Thus the same result is achieved as if a copy was made outside 28B. [No change.]

(k)  Bruce purchases a copy by download to his MP3 player, then lends the MP3 player to his sister.

OK.  28B does not come into play since no 'further copy' is made.  Private loan is unlikely to be an act restricted by copyright and, since the device itself is lent, no new copy is made.  See (f) above as to whether Bruce's sister would infringe when listening to the tracks.  [No change.]

The same would apply to a copy downloaded to any other device, including a storage device such as a USB stick.

(l)    Bruce purchases a copy by download to a USB stick, then sells the USB stick containing the original copy by online auction.

No, at least for the moment.  28B does not come into play since no 'further copy' is made.  Selling the USB stick infringes the copyright owner’s exclusive distribution right.  At present exhaustion of the distribution right applies only to physical copies (such as CDs) put into circulation by or with the consent of the rightsowner, not to a download made by a customer.  However it is possible that this may change in the future, should the CJEU expand its finding in UsedSoft so as to apply to non-computer programs.  This is a highly controversial issue. [No change.]

(m)    Bruce purchases a copy by download to a USB stick, makes a further copy from the download onto his hard drive for his private use, then sells the USB stick containing the original copy by online auction.

By making a further copy for his private use Bruce brings 28B into play.  Does that make the situation any different from where he has made no further copy at all? 

If Bruce does not destroy the further copy, then selling the USB stick by public sale is an infringement and, as in (j) above, the further copy will be treated as an infringing copy.  That is clear. 

What if Bruce does destroy the further copy?  Here things gets quite obscure, even controversial.   First, as explained above it may not be clear whether a download is a ‘lawfully acquired’ copy. [Now clarified to include a download.] If it is, then 28B(2)(b) states that copyright is infringed where an individual who has made a further copy within 28B permanently transfers the copy from which it is made, without destroying the further copy. 

The casual reader might think that it follows that if Bruce destroys the further copy, then transferring the copy from which it is made does not infringe copyright.  If 28B(2(b) does mean that, that would seem in effect to extend UsedSoft to non-software.  [No. See below.]]

Another view is that 28B provides a limited exception from infringement for certain further copies; that 28B(2)(b) describes a particular circumstance in which the exception is disapplied, and how in that circumstance the otherwise exempt further copy is to be treated as infringing (which it would have been if it had not fallen within 28B in the first place).  On that basis 28B(2)(b) says nothing about the underlying issue of whether transferring a downloaded copy of a non-computer program benefits from exhaustion of rights (as in UsedSoft) or not.  That may at some point come before the CJEU.  The section appears to accommodate either possibility. [The Government in its Response to the Technical Review states that 28B creates an exception only to the reproduction right and does not create any exception to the distribution right or the right of communication of works to the public. Nothing in those provisions creates a new right to re-sell copies, where that right does not already exist.]

The Good Fairy of the Internet and her Magic Wand

Last week I attended the Index on Censorship ‘Caught in the Web’ event.  I asked a question from the floor and was put on the spot by chair David Aaronovitch about David Cameron's call for the big players to do more about child porn on the internet.  Answered along the lines of not really understanding what more Cameron thought they could do, given that the Internet Watch Foundation runs a list already; and needing to examine any proposals to see if they carry too high a price in terms of interference with freedom of expression. 

OK as far as it went, but this is what I should have said.  When politicians demand that the most visible targets (ISPs, search engines, social media platforms) should do more to stop whatever abomination is hitting today’s headlines, really they just want a good fairy of the internet to wave a magic wand and make the nastiness go away.  The trouble is that the good fairy is a myth, the magic wand rubs out good stuff and the bigger the magic wand the more of the good stuff it rubs out.  And the nastiness still doesn’t go away.

Worst of all, when each has had its turn at waving the magic wand - the child protectionists, the responsible speakers, the cultural preservationists, the extreme speech purgers, the adcensors, the promoters of the public interest, the filth filterers, the religious reverents, the porn exterminators, the benevolent regulators, the upholders of community standards, the real namers, the guardians of the general good, the copyright cops, the internet safety inspectors, the privacy protectors, the departments of public harmony, the joke jailers, the moral champions, the speech cleaners and scrubbers and all the clamouring crew of prohibitionists – what then is left?

If the fundamental human right of free speech means anything, it is that we can build search engines without fear that the State will make us convert them into suppression engines.

I am not an IP address

[Update 3 December 2014. 18 months on, this has now come alive. The 'IP address resolution' proposals flagged up in the Queen's Speech are included in the Counter-Terrorism and Security Bill introduced in Parliament on 26 November. Initial analysis here. 

The Draft Communications Data Bill Joint Committee Report in December 2012 reached some consensus around this issue: "Not all United Kingdom providers currently obtain all the data necessary to trace which subscriber is using which IP address. During the course of our inquiry we heard of various circumstances in which the lack of this data has impeded investigations. We accept that if CSPs could be required to generate and retain information that would allow IP addresses to be matched to subscribers this would be of significant value to law enforcement. We do not think that IP address resolution raises particular privacy concerns."] 

The truism that an IP address denotes a device, not a human being, is ingrained in anyone with a technical understanding of the internet.  Nothing gets a geek going like the suggestion that an IP address identifies a person.  

So when the briefing document (PDF) on last week’s Queen’s Speech said: “When communicating over the Internet, people are allocated an Internet Protocol (IP) address”, sharp intakes of IT literate breath could be heard up and down the land.  The somewhat contradictory statement a few lines later that law enforcement had a problem with being unable to match individuals to IP addresses did little to improve matters, accompanied as it was by the suggestion that addressing this issue may involve legislation.

Cybergeeks are right to be exercised about this, a technical issue but no mere technicality.

Legislating around IP addresses is a dubious idea at the best of times.  A legislative approach that embeds specific technical implementations falls foul of an important principle of lawmaking in the tech field, technological neutrality.  Professor Chris Reed explains in his book Making Laws for Cyberspace how legislating at too detailed a technical level is likely to create legal uncertainty.  Legislation that assumes or mandates specific technology is at risk of impeding new technological developments.  Technology-specific law is bad law.

Legislating at a detailed technical level also requires a sound understanding of the technology.  The Digital Economy Act was the first UK statute to mention IP addresses.  One has to wonder whether that was done with a full appreciation of IP addressing's manyfold variations.

The particular problem on which the Queen's Speech focuses is that public IP addresses are often shared, so an IP address does not necessarily identify a single end user device.  The public IP address can denote the gateway of a domestic household or of a large organisation, a point within a public network or even the gateway of an entire public network (often mobile).  So thousands of household or organisation routers and potentially millions of end user devices may sit behind a single public IP address.  This is why, when the first draft Initial Obligations Code for the Digital Economy Act was published in May 2010, port numbers (not mentioned in the Act) were included as well as IP address information.    

Ironically, one thing that DEAct did appreciate was that an IP address does not identify a copyright infringer.  Much of the opposition to DEAct was aimed at its mechanism for overcoming this by creating a presumption, which the subscriber had to rebut, that a subscriber identified from the public IP address allocated by the subscriber's ISP was responsible for repeated infringement; even though anyone with access to the subscriber’s system could have used the device identified by the IP address. 

The courts know that this is an important issue.  The speculative invoicing business depended on asserting that a harvested public IP address provided grounds on which to assert a copyright claim against a subscriber identified from the IP address and the ISP's records.  In the UK this was discussed by H.H.J. Birss in the Media Cat case:

“IP addresses  are numerical references used to identify entities on the internet.

…

“Media CAT's monitoring exercise cannot and does not purport to identify the individual who actually did anything. All the IP address identifies is an internet connection, which is likely today to be a wireless home broadband router.  All Media CAT's monitoring can identify is the person who has the contract with their ISP to have internet access. Assuming a case in Media CAT's favour that the IP address is indeed linked to wholesale infringements of the copyright in question …, Media CAT do not know who did it and know that they do not know who did it.”

Numerous US courts have made the same point, summarised in May 2012 by magistrate judge Gary Brown in K-Beech:

“In sum, although the complaints state that IP addresses are assigned to “devices” and thus by discovering the individual associated with that IP address will reveal “defendants’ true identity,” this is unlikely to be the case. Most, if not all, of the IP addresses will actually reflect a wireless router or other networking device, meaning that while the ISPs will provide the name of its subscriber, the alleged infringer could be the subscriber, a member of his or her family, an employee, invitee, neighbor or interloper.”

Port numbers (if retained) can reach beyond the public IP endpoint to an end user device.  In combination with other contextual information it may then be possible to identify a person who was using the device.  For this reason law enforcement authorities want access to information associated with IP addresses and seek ways of matching an IP address to a unique end user device.  For exactly the same reason such schemes, even if embarked upon for the worthiest and most serious of motives, have civil liberties implications. 

These were articulated most starkly in the 1960s TV series The Prisoner.  When No. 6 said “I am not a number, I am a free man”, he did more than reject a numeric tag.  He pitted his personal autonomy against an all-controlling bureaucratic embrace: “I will not be pushed, filed, stamped, indexed, briefed, debriefed or numbered. My life is my own.”  Today Patrick McGoohan could as easily have written “I am not an IP address”.

The Convergence Green Paper that the European Commission didn’t write

The new European Commission Green Paper “Preparing for a Fully Converged Audiovisual World: Growth, Creation and Values” seems to have been a mouthful even for the Commission’s own PR department.  The press release translates it into ‘Internet on TV, TV on Internet’ - which does give a better flavour of what the Green Paper is about.

The Green Paper is a typical Commission mixture of current technology trends (dual screening gets a mention), market statistics leavened with a few anecdotes, lobbyists’ issues of the moment and hints of interventions that the Commission may have in mind for the future.  These are all underpinned by an unwavering belief in the efficacy of State-led intervention to achieve the Commission’s chosen public policy objectives.

As is de rigeur in documents such as this, the Commission has a grand vision:

“to seize the opportunity of this changing technological environment to ensure the widest possible access to European diversified content for all Europeans and, the widest choice of high quality offers”. 

That hints of a more interventionist Fortress Europe attitude than the Commission’s December 1997 Convergence Green Paper (PDF):

“This first step is intended to pave the way for the development of an appropriate regulatory environment which will facilitate the full achievement of the opportunities offered by the Information Society, in the interests of Europe and its citizens as the 21st century begins.”

Beyond the grand vision there is some fairly controversial stuff in the new Green Paper, particularly around the possibility of extending the Audiovisual Media Services Directive both in scope and to non-EU service providers. 

But I am afraid I tired of the DG-speak.  I started to imagine the 2013 Convergence Green Paper that the Commission could have written, but didn’t. 

Here it is.

“15 years ago, in December 1997, we at the Commission published a Convergence Green Paper.   We said:

“… the Commission's Communication on electronic commerce … proposed the principle of 'no regulation for regulation's sake'.   This principle applies equally to all areas of convergence.”

We tentatively suggested five key regulatory principles that should govern our activity in the field of convergence.  They included:

“Regulation should be limited to what is strictly necessary to achieve clearly identified objectives.”

Of those five regulatory principles that we suggested in 1997, this remains the most important.  We observed then that:

“public authorities must avoid approaches which lead to over-regulation, or which simply seek to extend existing rules in the telecommunications and media sectors to areas and activities which are largely unregulated today.”

The Commission recognises that the pursuit by government of general policy goals – such as media pluralism, cultural diversity, protection of consumers and minors, media literacy – carries with it the potential for unintended consequences and regulatory failure and to undo the general good that comes from freedom of private action.  The Commission appreciates that government failure is a more significant risk than market failure.    

Adhering to the principle of strict necessity, the Commission’s goal now must be to clear away all regulatory and legislative obstacles to future innovation in the audiovisual sphere.  The only permissible exception can be regulation that is conclusively demonstrated, by the most stringent criteria, to be both necessary and effective for the achievement of a generally agreed legitimate public policy objective; which carries a demonstrably minimal risk of unintended or damaging negative consequences, risk of regulatory capture or other regulatory failure; and which does not disproportionately interfere in fundamental human rights, in particular the right of freedom of speech.

The Commission recognises that the internet has brought forth the greatest flowering of individual communication and access to knowledge since the dawn of mankind.  Speech on the internet is a delicate flower which must not be damaged.  The Commission applauds the comments in the Joint Dissenting Opinion in the recent Animal Defenders case in the European Court of Human Rights:

“there seems to be an inherent contradiction in a viable democracy safeguarded by broadcasting restrictions. … A robust democracy is not helped by well-intentioned paternalism.”

Stated simply, the Commission’s role is now to clear the field of regulation and move out of the way.  This will enable the peoples of Europe to make their own viewing and listening choices at all levels: international, nation, group, community and individual.  If some people prefer Disney to Hugo, that is an expression of their individual autonomy and right to cultural self-determination.  As such it is outside the legitimate sphere of action of the Commission.

Each person in Europe must be free to choose and pursue their own personal goals.  It is not for the Commission to formulate policy goals and then to expect private actors, economic or otherwise, to pursue objectives that we set for them.  The Commission does not know what form any future innovation will take, or what the pace of innovation will be.  Nor can it ever know.  It is not the Commission’s place to attempt predictions, still less to formulate interventions based upon them. 

The Commission does not believe that it should set concrete targets to be achieved by some arbitrary future date.  We at the Commission shiver with embarrassment at the memory of the programme to ‘Complete the European Single Market by 31 December 1992’ - for all the world like some Soviet era tractor production plan.

The Commission recognises that innovation brings change.  It is no part of the Commission’s role to save legacy actors from the consequences of change, nor to try to influence future change in a particular direction.  The Commission recognises that any adverse consequences of change for particular sectoral and vested interests are outside the Commission’s legitimate sphere of action.  The Commission strongly believes that the European taxpayer should not to be compelled to subsidise historic industries in the name of cultural diversity. 

Turning specifically to the audio-visual sector, in the 1997 Green Paper the Commission said:

"… convergence may lead to less regulation in telecommunications and media sectors, and should not lead to more regulation in areas such as IT."

And

"Convergence may challenge current regulatory approaches, particularly, with regard to licensing of networks and allocation of resources, where such approaches reflect a perceived scarcity of both radio-frequency and of content."

"…in a fully digital environment, scarcity may over time become a less significant issue, calling for current regulatory approaches to be reassessed."  

Since scarcity has now all but disappeared, other than where artificially created by regulation, the time has come to reassess current regulatory approaches. 

Ideally the Commission would propose to sweep away the archaic structures of broadcast regulation, which have their roots in the government spectrum grabs of the mid 20^th century.  However we recognise that most Member States will continue to cling to these regulatory models like the man with the red flag in front of the early motor car. 

Pragmatically, broadcast regulation is likely to wither as it becomes less and less relevant.  It is however imperative, as articulated in the 1997 Green Paper, to prevent outdated broadcast regulatory models from spreading any further, especially to the internet.  We now realise that it was a serious error in the AVMS Directive to extend regulation to so-called ‘TV-like’ on-demand services.  We will roll that back at the earliest opportunity.

The cross-border aspect of the internet is becoming increasingly significant.  The AVMS Directive does not apply to services originating outside the EU.  The Commission recognises that the people of Europe are readily able to understand that incoming foreign services are created under different sets of legal standards and to appreciate those services in that context. 

In the spirit of international cultural diversity the Commission believes that the fewest possible obstacles should be put in the way of incoming non-EU services, including where they are targeted at the EU.  The Commission would regard it as an act of unwarranted arrogance for the EU to seek to apply its content laws to services originating in non-EU countries and regions. 

The choice of European citizens to access and be informed about non-EU services that may not adhere to European content laws or standards should therefore not be denied or hindered by any Member State.  The Commission especially welcomes the opportunity for European citizens to share in the rich and diverse cultural heritage of the USA.”

Does copyright control browsing? Meltwater in the UK Supreme Court.

[Updated with CJEU judgment of 5 June 2014]
Back in July 2011 I commented on the Court of Appeal judgment in NLA v Meltwater and explained how the reach of digital copyright had accidentally been increased beyond that in the offline world.  That was as a result of accepting that transient and temporary copies created in computer memory count as copies for copyright purposes.  This potentially extended copyright control to all user activities involving a computer, including browsing the web. 

The Public Relations Consultants Association subsequently appealed the Meltwater decision to the UK Supreme Court, which has now referred to the EU Court of Justice the question of whether browsing a website requires the permission of the copyright owner.

Technically, the question is whether browsing falls within the 'temporary copies' exception of Article 5.1 of the EU Copyright in the Information Society Directive.  Lord Sumption, who gave the judgment of the UKSC, concluded that browsing did fall within the exception, and so did not require the permission of the copyright owner.  However he recognised that the issue has a transnational dimension and has important implications for many millions of people across the EU making use of the internet.  That made it desirable that the decision be referred to the CJEU, so that the critical point might be resolved in a manner that would apply uniformly across the EU.

From a broader perspective the judgment is of particular interest for the extent to which the Supreme Court recognised the disparity, if browsing does require the copyright owner's permission, between the reach of copyright offline and online.  The judgment opens by pointing out that:

"Broadly speaking, it is an infringement to make or distribute copies or adaptations of a protected work. Merely viewing or reading it is not an infringement. A person who reads a pirated copy of a protected book or views a forgery of a protected painting commits no infringement although the person who sold him the book or forged the painting may do."

It goes on to explain that in the course of the ordinary use of the internet:

"[the user's] object is to view the material. The copies temporarily retained on the screen or the internet cache are merely the incidental consequence of his use of a computer to do that. The question which arises on this appeal is whether they are nonetheless infringing copies unless licensed by the rights owner."

Lord Sumption observed that the question:

"potentially affects millions of non-commercial users of the internet who may, no doubt unwittingly, be incurring civil liability by viewing copyright material on the internet without the authority of the rights owner, for example because it has been unlawfully uploaded by a third party. Similar issues arise when viewers watch a broadcast on a digital television or a subscription television programme via a set-top box."

Having analysed the relevant EU caselaw and concluded that browsing fell within the temporary copies exception, Lord Sumption addressed the consequences of his conclusion.  He said:

"We were pressed with the argument that if the viewing of copyright material on a web-page did not require a licence from the copyright owner, he would be exposed to large-scale piracy of a kind which would be difficult to detect or prevent.

I am not persuaded by this argument. … it has never been an infringement, in either English or EU law, for a person merely to view or read an infringing article in physical form. This state of affairs, which is recognised in the enumeration of the copyright owner's rights in articles 2, 3 and 4 of the Directive, has never been thought inconsistent with a high level of protection for intellectual property.

All that article 5.1 of the Directive achieves is to treat the viewing of copyright material on the internet in the same way as its viewing in physical form, notwithstanding that the technical processes involved incidentally include the making of temporary copies within the electronic equipment employed.

… [I]f it is an infringement merely to view copyright material, without downloading or printing out, then those who browse the internet are likely unintentionally to incur civil liability, at least in principle, by merely coming upon a web-page containing copyright material in the course of browsing. This seems an unacceptable result, which would make infringers of many millions of ordinary users of the internet across the EU who use browsers and search engines for private as well as commercial purposes."

…  [I]f merely viewing a web-page is not an infringement, that does not leave the copyright owner without effective remedies against pirates. It simply means that his remedy must be found against others who on the face of it are more obviously at fault. Nothing in article 5.1 impairs the copyright owner's right to proceed against those who unlawfully upload copyright material onto the internet, just as the copyright owner has always been entitled to proceed against those who make or distribute pirated copies of books, films, music or other protected works."

These passages exhibit a strong policy interest in achieving functional equivalence: for copyright to apply to the internet as far as possible in the same way that it did to physical copies before the digital revolution.  In that sense the Supreme Court judgment has similarities to the Canadian Supreme Court Copyright Pentalogy and the CJEU decision in Usedsoft v Oracle.

The approach articulated by Lord Sumption is the opposite of that stated by Proudman J at first instance: "The exception cannot have been intended to legitimise all copies made in the course of browsing or users would be permitted to watch pirated films and listen to pirated music." Lord Sumption, albeit with the benefit of later CJEU decisions in FAPL v QC Leisure and Infopaq II that were not available to Proudman J or the Court of Appeal, has made a clear statement that there is no policy reason why mere watching and listening (as opposed to downloading or printing) should be regarded as an infringement in the online world any more than in the offline world.  We now await finding out whether the CJEU agrees. [Yes, it does.]

Ten ways in which copyright engages freedom of expression

"If the protection of expression is itself a restriction on freedom of expression, then we have entered a metaphysical Wonderland that stands logic on its head, and undermines core, shared global values about personhood." Recording Industry Association of America, October 2014.

"... some restriction on expression is the inherent and intended effect of every grant of copyright". US Supreme Court, Golan v Holder, January 2012. 

Reader Alert: this is a very long (3,750 word) post.  If you would prefer to cut to the chase, here are my ten axes along which copyright and freedom of expression intersect:

1.      Expression to Ideas

2.      Expression to Opinion

3.      Originality to Facts

4.      Dissemination to Reference

5.      Zero to Perpetuity

6.      Suppliers to Users

7.      Participation to Facilitation

8.      Copying to Communicating

9.      Pennies to Prison

10.   Targeted to Scattergun

Slideshow presentation to ORGCon 2013 based on this post:

Who wins when copyright and free speech clash? from Graham Smith

The recent Donald Ashby (sub nom Ashby Donald) decision of the European Court of Human Rights has revived interest in the relationship between copyright and freedom of expression. The litigation arose because two of the defendant photographers had put on their US website pictures taken by the third at the Paris fashion shows. Under French law that requires the permission of the couture houses. The highest French Court had confirmed their conviction for criminal copyright infringement.

The photographers lost their appeal to the Human Rights Court. But despite the loss the Court’s decision does finally explode any residual myth that copyright has no impact on freedom of expression.

Article 10 of the European Convention on Human Rights protects freedom of expression as a qualified right. In Ashby the Court explicitly held that the appellants’ activities fell within the exercise of the right of freedom of expression and that the conviction interfered with that. It was therefore necessary to consider whether the interference was justified. However the Court held that in this particular case it could not find that either the photographers’ conviction or the financial penalties were a disproportionate interference. Particular account had to be taken of the Contracting State’s broad margins of appreciation for both commercial speech and the protection of rights of others. 

In the subsequent case of Neije and Sunde the ECtHR found that the appellants' criminal convictions for operating the Pirate Bay site interfered with their right to freedom of expression, since they had put in place the means for others to impart and receive information within the meaning of Article 10 of the Convention.  In this case the Court found that there were weighty reasons for the interference and held the complaint to be manifestly ill-founded and inadmissible.

Ashby provides a good opportunity to review the various ways in which copyright and freedom of expression can interact. Some have asserted that copyright and freedom of expression exist on different planes which do not intersect at all. In 2000 Mance L.J. said in Hyde Park Residence Ltd v Yelland: “Copyright does not lie on the same continuum as, nor is it the antithesis of, freedom of expression.”

18 months later Lord Phillips, giving the judgment of the whole Court of Appeal in Ashdown v Telegraph Group Ltd, said the opposite: “Thus copyright is antithetical to freedom of expression. It prevents all, save the owner of the copyright, from expressing information in the form of the literary work protected by the copyright.”

The US Supreme Court in Golan v Holder has recognised that “some restriction on expression is the inherent and intended effect of every grant of copyright.”

Some have argued that, even accepting that copyright does interfere with freedom of expression, the necessary balance between that and other legitimate interests is struck entirely internally within copyright legislation. Copyright is therefore not susceptible to external human rights scrutiny.

In England that argument was put to rest in Ashdown. If any doubt remained at a European human rights level, Ashby has put paid to it. As a matter of EU law, the CJEU in SABAM v Scarlet has stated that intellectual property rights are neither inviolable nor must be absolutely protected, but must be balanced against other fundamental rights including freedom of expression.

Ashby on its facts was an obvious interference with the photographers’ Article 10 rights, even if the ECtHR was unable to find the interference disproportionate. The defendants had put the photographs of the Paris catwalk shows on their public website. The fashion houses were using copyright to try to limit and control public dissemination of images of the shows. The engagement with the right to receive and impart information is clear.

Copyright can engage freedom of expression more subtly. Depending on the scope of copyright and the nature of the remedies against infringers (or in some cases even against non-infringers), when applied to any particular set of facts there are many ways in which copyright can step over the line and disproportionately interfere with Article 10 rights.

It is difficult, perhaps impossible, to identify clear dividing lines between proportionate and disproportionate interference. This is partly because copyright is itself regarded, at least in European human rights law, as a property right that has to be weighed in a balancing exercise with other rights such as freedom of expression. It is also because the degree of interference may result not just from one aspect of copyright, but from the combination of several; and because the impact depends on the factual circumstances of each individual case.

Different features of copyright can engage freedom of expression on different axes. In this sense copyright can perhaps be likened to a mixing console, on which each aspect of copyright that potentially engages freedom of expression is represented by a slide control. One control set to maximum may itself be disproportionate, either per se or in its application to a particular set of facts. Equally the combined effect of several controls set to less than maximum could also be disproportionate.

Take the example of an “originality” slide control. If the slider were set at a level that restricted the use of high level ideas as well as detailed expression, then the interference with the Article 10 right might be disproportionate in itself; but even if not, the interference would be that much the greater if the remedy for infringement were an injunction preventing publication as opposed to a small amount of monetary compensation that permitted continuing dissemination.

Rather than attempt to identify a clear line between proportionality and disproportionality, or examine how rights should be balanced against each other, this article aims to do no more than describe the mixing console: identify each slider whose movement up or down the scale is capable of engaging Article 10 rights to a varying degree.

The focus of this article is on engagement with freedom of expression and nothing else. Arguments of all kinds can be deployed for and against particular aspects of copyright. No given argument necessarily has anything to do with freedom of expression. If it does not, it won’t feature here.

We should not forget some less prominent creatures of statute such as legislation prohibiting circumvention of copyright protection measures, and EU database right (which is turning out in some respects to be more restrictive than copyright). However for the sake of simplicity this article is limited for the most part to copyright. It is written largely from a UK perspective. Copyright in other countries may differ.

The copyright and freedom of expression mixing console

Article 19 of the Universal Declaration of Human Rights states:

“Everyone has the right to freedom of opinion and expression; this right includes freedom to hold opinions without interference and to seek, receive and impart information and ideas through any media and regardless of frontiers.”

In short, the fundamental right of free speech relates to the free flow of ideas, opinion and information.

Our first group of slide controls concerns the extent of copyright and its interaction with each of these three aspects of freedom of speech.

Slider 1: From Expression to Ideas

It is a truism, if not always an illuminating one, that copyright protects expression and not ideas. Jacob J. (as he then was) cautioned in IBCOS Computers Limited v Barclays Mercantile Highland Finance Limited [1994] FSR 275 that while merely taking a sufficiently general idea does not infringe, to take a detailed “idea” may do so. It is a question of degree.

Wherever and however fuzzily this line may be drawn, one reason why it exists is the recognition that all expression, however original, owes something to what has gone before –and the more general the previous knowledge the more likely it is to do so. To insist that all expression should embody only completely original ideas, shorn of anything that has gone before, would be to require silence. If we were to push the expression/idea copyright slider too far, the babble of voices would gradually diminish to nothing. Well of course it wouldn’t, since such a law would lose all legitimacy long before silence fell. But the serious point is that the idea/expression slide control engages freedom of expression more intensely towards the ideas end of the scale.

The proposition that the idea/expression dichotomy engages freedom of expression was recognised in, for instance, the US Supreme Court decision in Eldred v Ashcroft:

“… copyright law contains built-in First Amendment accommodations. … First, it distinguishes between ideas and expression and makes only the latter eligible for copyright protection. … As we said in Harper & Row, this idea/expression dichotomy strike[s] a definitional balance between the First Amendment and the Copyright Act by permitting free communication of facts while still protecting an author’s expression. … Due to this distinction, every idea, theory, and fact in a copyrighted work becomes instantly available for public exploitation at the moment of publication.”

At the other end of the scale it may be said that if expression is protected too narrowly, then creativity is inhibited through lack of sufficient incentive to multiply its fruits. This is a common enough argument. Copyright was described in the Harper & Row decision of the US Supreme Court as ‘the engine of free expression’. This argument engages a fundamental right of freedom of expression only if one regards the State as having a positive obligation to promote freedom of expression, as opposed to simply refraining from interference with it.

Slider 2: From Expression to Opinion

The right of freedom of expression is most robustly protected by international human rights law when opinion, especially political opinion, is at stake. So European human rights law has developed a hierarchy in which individual political expression is more robustly protected than, say, commercial speech. As reiterated in Ashby, the ECHR allows Contracting States a greater margin of appreciation (i.e. allows them more latitude to interfere) with commercial speech.

The potential for copyright to interfere on the opinion axis is high, even when an element of expression is taken. Hence the existence of various exceptions for activities such as criticism and review and for parody. These define circumstances and conditions in and under which it is legitimate to take some element of expression. However, specific exceptions are not always adequate. In the USA the more flexible fair use defence is recognised as an essential part of copyright’s accommodation with the First Amendment:

“We then described the “traditional contours” of copyright protection, i.e., the “idea/expression dichotomy” and the “fair use” defense. Both are recognized in our jurisprudence as “built-in First Amendment accommodations.” (Golan v Holder)

In Europe, for political expression direct recourse has sometimes successfully been made to external Article 10 considerations. In the Netherlands the Nadia Plesner Darfurnica case, although it concerned Community design right rather than copyright, is an example of freedom of political expression trumping the exercise of an intellectual property right.

Slider 3: From Originality to Facts

Works of authorship – literary, artistic, dramatic and musical works - attract copyright only if they are original.

Originality is closely related to infringement. For instance the infringement test as formulated by the EU Court of Justice in Infopaq is whether a part of the author’s own intellectual creation has been taken.

So the level at which the threshold for originality is set affects not only whether a work is protected at all by copyright, but what kind of use of the work the copyright owner can prevent. The degree of use that can be prevented is also affected by whether the infringement threshold is, for instance, the taking of any part or the taking of a substantial part.

Where the work contains factual information, a low originality threshold can protect not just the author’s form of expression, but the facts themselves. Facts are at the most basic level of information that humans may wish to impart or receive. The extract from Eldred v Ashcroft quoted above makes clear the engagement with freedom of expression.

A low originality threshold may overly restrict the ability to recommunicate facts imparted through the means of a copyright work. The potential for this consequence is why exceptions exist such as fair dealing for the purpose of reporting current events. These may be quite limited. For instance in UK law the exception does not apply to photographs.

Beyond copyright the ability to recommunicate facts is increasingly restricted by database right, which is specifically aimed at protecting otherwise unprotectable collections of data. Database right has few exceptions, an investment rather than an originality threshold and (as currently interpreted by the courts) hair trigger infringement criteria.

Slider 4: From Dissemination to Reference

Copyright has always sought to prevent the dissemination of infringing copies. It has typically stopped short of preventing people from referring to the existence of infringing material, whether in a library index, in a footnote to an article, in a newspaper or whatever.

When copyright, in its enthusiasm to prevent the spread of infringing copies, seeks to prevent people from acknowledging their existence or pointing to where they may be found, it engages the right of freedom of speech.

In the digital and online world it does so most clearly when it seeks to characterise linking and similar acts as infringement, or bears on the creation and use of search engines. For activities such as image or video search the engagement may take on further aspects, for instance if effective referencing depends on the ability to make and store a thumbnail of the image or of a frame of video.

Slider 5 From Zero to Perpetuity

It is difficult to identify any particular duration of copyright that triggers engagement with freedom of expression; although if it is accepted that any copyright protection necessarily affects freedom of expression, then the longer the duration the greater the interference. The longest durations of copyright tend to give rise to significant orphan works issues - an especially visible engagement with the freedom to receive information.

Attempts in the USA to challenge copyright extensions on First Amendment grounds have not fared well. In Eldred v Ashcroft term extension to life plus 70 years involving resurrection of expired copyright for some works was held not to offend against the First Amendment. In Golan v Holder introduction of copyright for foreign works previously in the public domain was held not to violate the First Amendment.

Our second group of slide controls concerns who infringes and by doing what.

Slider 6 From Suppliers to Users

The wider the net of infringement is thrown, the greater the scope for copyright to engage with freedom of expression.

Traditionally copyright tended to concentrate on plagiarists or creators who ‘overborrowed’ from the work of others and on manufacturers and distributors.

Copyright bit hard on those at the head of the duplication supply chain: the owners and operators of printing presses, vinyl pressing plants, CD manufactories. It bit less hard on the middlemen between them and the final purchaser. In general they infringed only if they knew they were dealing with infringing copies.

At the point of retail sale it was not an infringement to purchase an infringing item. Even less was it an infringement to read or view an infringing item. Pre-digital, the burden of determining what was and was not infringing was thought not appropriate to place on the user of copyright works.

Digital technology has changed all that. The ubiquity of temporary and transient copies within user devices has provided the opportunity to argue that mere enjoyment of copyright works can infringe, and on a strict liability basis at that.

Enjoyment as infringement is in broad terms the issue currently before the UK Supreme Court in Meltwater (sub nom Public Relations Consultants Association Ltd v The Newspaper Licensing Agency Limited and others). It is a telling indication of how far things have moved in the digital era that the judge at first instance in that case was able to say that the temporary copies exception “cannot have been intended to legitimise all copies made in the course of browsing or users would be permitted to watch pirated films and listen to pirated music.” (emphasis added)

Users and consumers are the least well equipped to determine whether something is protected by copyright, whether the purpose for which it is to be used requires authorisation, and if so whether it is authorised or not. In human rights terms, placing such a liability burden on the user may chill the right of freedom of expression. since the user may be deterred from legitimately accessing information and knowledge.

Slider 7 From Participation to Facilitation

This issue concerns the degree of participation in an infringing activity necessary to be regarded as an infringer. Imposing liability on those who volitionally do the act or cause it to be done is (questions of strict versus fault-based liability apart) uncontroversial. But if liability is extended to those who merely facilitate or enable an infringement, that is likely to have an impact on legitimate activities which otherwise promote the dissemination of lawful material. The right of freedom of expression is consequently engaged.

Similar issues arise where accessory liability rules (such as those of joint tortfeasorship) impose liability on one person for the infringing act of someone else. The English courts have tended to draw a reasonably clear line between facilitation or assistance (non-liable) and involvement of a kind that makes the other person’s infringement one’s own (liable).

A particular area of potential conflict concerns online intermediaries. Conduits, hosts and other similar entities are the engines that drive the free flow of information on the internet. As platforms and pipes they can be regarded as facilitators. Many seek to characterise them as profiting from infringement and duty bound to prevent the use of their services by infringers. Regardless of the rights and wrongs of that debate, it has long been recognised that placing liability burdens on intermediaries has the potential to interfere with freedom of expression. This was specifically recognised in Recitals (9) and (46) of the Electronic Commerce Directive, which provided liability shields for conduits, caches and hosts.

Potential, as much as actual, risk of liability is capable of engaging with freedom of expression if a liability regime incentivises the removal of material that may not in fact infringe. This is often raised as an issue with ‘notice and takedown’ regimes.

Slider 8 From Copying to Communicating

It is natural to think that copyright is about copying. Indeed it is, but not only so. A whole range of acts restricted by copyright now exists. Some types of restricted act are more prone directly to engage freedom of expression than others. One of the most obvious is communication to the public, especially when deployed against linking.

Sir Tim Berners Lee said in the early days of the Web:

“The first amendment to the Constitution of the United States … addresses the right to speak. The right to make reference to something is inherent in that right. On the web, to make reference without making a link is possible but ineffective - like speaking but with a paper bag over your head.”.

The question of linking as infringement is a live issue. It is the subject of the current Svensson reference to the CJEU, in which some rightsowners are asserting a right to prevent others from linking without permission to authorised copyright material on their own sites.

The right of freedom of expression is also engaged in the context of links to unauthorised material. Cases such as Newzbin in the UK have held that at least in certain behavioural contexts some kinds of linking can amount to infringement by unauthorised communication to the public of the linked-to material.

The third group of sliders concerns Remedies

The degree of interference with freedom of expression is not just a question of liability, but also of remedies. The appellants in Ashby complained separately about the conviction for copyright infringement and the amount of the financial penalties.

Slider 9 From Pennies to Prison

The nature of any remedy enforcing an aspect of copyright that engages freedom of speech will bear on the proportionality of the interference with freedom of speech. Here we simply list some relevant factors in assessing the proportionality of remedies against infringers:

- Criminal versus civil liability.

- For criminal liability: imprisonment versus fines.

- Compensatory awards versus aggravated, exemplary or punitive damages.

- Compensatory awards versus fixed (statutory) damages.

- Injunctions against future dissemination and the penalties for breach of an injunction.

Slider 10 From Targeted to Scattergun

Remedies that are not limited to the dissemination of infringing material, but have the potential to interfere with access to legitimate non-infringing material or activities, engage the right of freedom of expression in a particularly direct manner. Suspension or termination of internet access is one obvious example.

The engagement is more apparent when remedies are made available against parties such as intermediaries who do not infringe or, if the law does provide that they infringe, for whom freedom of expression is acutely engaged as a result of the nature of their activities (see Slider 7).

The CJEU in SABAM v Scarlet held that broad indefinite filtering injunctions against internet access providers (and against hosts in SABAM v Netlog) were incompatible with the freedom of expression rights of users of the services:

“that injunction could potentially undermine freedom of information since that system might not distinguish adequately between unlawful content and lawful content, with the result that its introduction could lead to the blocking of lawful communications.

…

Consequently, it must be held that, in adopting the injunction requiring the ISP to install the contested filtering system, the national court concerned would not be respecting the requirement that a fair balance be struck between the right to intellectual property, on the one hand, and the freedom to conduct business, the right to protection of personal data and the freedom to receive or impart information, on the other."

The UK courts in cases such as Newzbin2, Dramatico v BSkyB and EMI v BSkyB have recognised that blocking injunctions against ISPs engage the freedom of expression rights of internet users.

Where the site requested to be blocked is unrepresented before the court and the defendant ISPs do not contest the injunction, procedural safeguards may be relevant to ensure that the court is fully equipped to evaluate the implications for freedom of expression of internet users.