Regulatory convergence - same old question, same right answer

Browsing through the responses to the Communications Review Open Letter in advance of the forthcoming Green Paper, I came across the answer from the Oxford Internet Institute to the posed question "Is regulatory convergence across different platforms desirable and, if so, what are the potential issues to implementation?".    The OII says, very sensibly:

"... it would be a significant mistake to seek regulatory convergence across platforms if this means imposing a model of broadcast regulation on the Internet. It is often assumed that the Internet is a modern era ‘Wild West’, lawless and unregulated. In fact, the opposite is true – there is already extensive regulation of Internet service provision, content and activities. We would argue that traditional regulatory models for broadcasting, common carriers (such as post or telecommunications) and the press cannot be imposed wholesale on the Internet without serious risks to its vitality and its contribution to the UK economy as well as potential chilling effects of speech."

I was reminded of another Green Paper 14 years ago - the European Commission Convergence Green Paper - which raised much the same question.  At the time, alarmed at the possibility that policymakers might unthinkingly apply anachronistic broadcast regulation to the internet, I wrote a piece for the Financial Times (8 April 1998) entitled "Networks without broadcast constraints".  Unblushingly I reproduce it below, if only to suggest that while the question may recur, the right answer remains unchanged. 

“In the 17th century John Milton, in his celebrated Areopagitica, argued against Oliver Cromwell's licensing of books and newspapers. Today,with the rise of the internet, the battle has begun for freedom of the digital networks.

The battleground is Brussels. Last December the European Commission published a Green Paper called "Convergence of the Telecommunications, Media and Information Technology Sectors, and the Implications for Regulation".

Convergence is the phenomenon that, thanks to advances in IT and communications technology, content no longer respects the technical boundaries of telephone, radio, television and cable. The Green Paper identifies the internet as "both the symbolic and prime driver". The Green Paper, open for consultation until the end of April, begins a debate about the regulatory environment for digital content. The outcome is important for traditional print media. As paper content migrates to digital networks, should digital content be licensed and subject to discretionary regulation as if it were broadcast?

Or should it be like print, which in liberal countries needs no licence and whose publishers are answerable only to general laws enacted by democratic institutions and enforced by independent courts ? In the rush to devise policy proposals for the internet, it is easy to assume that it is another variety of new media, which like broadcast should subject to discretionary content regulation.  Even if that could be justified, the analogy is misconceived. Spectrum on the internet is far from scarce. Internet content is not invasive. Every internet user is also potentially a publisher. The US Supreme Court, in the Communications Decency Act case, accepted these differences from broadcast media. It concluded that the net's content is as diverse as human thought and it deserved the highest degree of First Amendment protection.

Even so, digital networks remain at risk of having broadcast regulation imposed upon them. Outside Europe, Singapore has imposed a class licensing scheme for the internet under the control of the Singapore Broadcasting Authority. The UK Broadcasting Act arguably could apply to moving pictures on the net.

If the Broadcasting Act were to be extended to web versions of UK newspapers, they would have to be censored to comply with the Act's requirements of political and public policy neutrality. No editorials on the web. Freedom of speech would be the preserve of an ever-diminishing print rump. The Green Paper does acknowledge the principle of "no regulation for regulation's sake". It rightly points out that the IT industry operates in an unlicensed environment. It mentions that convergence may challenge licensing approaches based on perceived scarcity of radio-frequency and content. It refers to the possibility that different standards may apply to the same content provided over different platforms. It tends to favour lighter regulation. However, as paper content migrates to the networks, the Green Paper pays little attention to the need to preserve the freedom to publish that content without permission from the state.

Nowhere does it suggest a clear principle that licensing and discretionary regulation are inappropriate for content published over networks. Instead the paper tends to emphasise "self-regulatory" approaches that risk verging on government regulation by another name.

It points out that self-regulation for the net may lead to divergent approaches unless co-ordinated to some degree at a community level. So much for diversity - the essence of the internet - as a positive good. Martin Bangemann, EU commissioner, has called for an International Charter for global communications. He also suggested the possible need for a European Communications Act covering infrastructure, services, content and conditions of access.

A recurring theme in the debate is the need to regulate internet content effectively. Mr Bangemann, describing reactions to his call for an international charter, says that some people who are afraid we are heading for anarchy considered the internet something that must be brought under strict control.

The real need is to free the internet and other digital networks from the threat of censorship by licensing and discretionary regulation, and to establish them as subject only to settled laws protecting free speech, enforced by independent judges. If that were embodied in Mr. Bangemann's charter, it would be progress indeed. Achieving it in a world replete with countries whose governments do not value freedom of speech should be the important task for international lawmakers.”

Avoidable ATVOD

Following OFCOM’s ruling in the Sun Video case, what does the future hold for ATVOD? 

ATVOD sees itself as a model for co-regulation, whose expertise can usefully inform the forthcoming Communications Green Paper: 

“Our experience is that co-regulation of video on demand services has proved capable of yielding nimble, economical solutions and the promise of establishing a broad consensus around light touch regulation.” (Letter to Minister for Culture, Media and Sport, 28 June 2011)

Others view it differently.  In July 2011 Lord Clement-Jones criticised ATVOD in Parliament as expensive, too wide in scope, far from light touch and already giving rise to litigation. 

An answer to a Parliamentary question in November 2011 revealed that complaints to ATVOD about audiovisual services were running at about two a day since April 2011.  No complaints had resulted in a finding of breach of content standards.  ATVOD costs around £450,000 per annum, recouped by fees levied on about 150 regulated VOD services.

The question is bound to be asked, come the Communications Green Paper, whether ATVOD should now be given a decent burial.  What purpose is served by an extra layer of content regulation - whether ATVOD-style co-regulation or a full-blown statutory regulator – over and above the general law, especially when funded by imposing substantial costs on a small section of industry?

Are there alternatives?  The UK government does have to comply with the AVMS Directive, which lays down content requirements specific to TV-like audiovisual services.  However those can be enshrined in a few paragraphs of statute, with a sanction such as the ability for a person affected to apply to court for an injunction.  That, in conjunction with a voluntary code of conduct, is how the Irish government has implemented the AVMS Directive.

Subjecting on-demand audiovisual services to an appropriately crafted statute would remove the need for a funded regulatory or co-regulatory body and provide a regime much closer to that applicable to most other speech and content, both generally and on the internet.

An Irish-style approach was never considered when the UK government consulted on the implementation of the AVMS Directive in 2008.  The Communications Green Paper would provide the perfect opportunity to reconsider the need for any regulator at all. 

OFCOM reins in VOD regulator’s reach

ATVOD’s high-profile campaign to bring video sections of newspaper websites under its regulatory wing has hit the buffers. 

Just before Christmas OFCOM ruled, in an appeal from an ATVOD determination, that the Sun Video elements of the Sun website are not a TV-like video on demand service.  ATVOD has promptly dropped similar cases against eight other newspaper and magazine sites.

OFCOM found that ATVOD’s original determination on the Sun site concentrated too much on the part of the in which the video links were collected together, rather than the site as a whole, and were wrong to conclude that the video links were a separate service in their own right.  They were an incidental part of the Sun newspaper site, and it was not possible to identify anything within the site whose principal purpose was the provision of audiovisual material.  OFCOM having reached this conclusion, it did not go on to consider whether the audiovisual material was comparable to a television service.  

OFCOM did address the general question of whether the Sun’s website provided a service that the Directive sought to bring within its regulatory scope.  In particular OFCOM considered Recital 24: whether the nature and the means of access to the service would lead the user reasonably to expect regulatory protection within the scope of this Directive.

OFCOM concluded that the user would not regard the audiovisual material on the Sun website as competing with television programmes, nor expect what he was viewing to be regulated as such. Rather, as far as the written contents of the website and the Video section were concerned, the user would likely regard himself as viewing the electronic version of The Sun newspaper.

The 47 page OFCOM decision gives detailed guidance on the factors that are likely to influence whether provision of audiovisual material is a principal purpose of a website.  These include:

-         Whether the audiovisual material has its own homepage; or by contrast whether it is accessible via a homepage that is styled as providing, and in practice does provide, some other service with its own independent identity

-         Whether, where it is made available on a website providing other content, such as written articles, a significant amount of the relevant audiovisual material is catalogued and accessed via a separate section of the website

-         Whether the audio visual material is presented or styled and marketed as a television channel, e.g. as ‘X TV’

-         Whether a significant amount of the audiovisual material is of a substantial duration and/or comprises complete programmes, rather than bite-sized clips or extracts from longer programmes; and whether it has an independence that means it is watched and fully understood on its own.  Material comprising clips of short duration whose context, meaning and significance is only properly or fully understood by reading accompanying written material is more likely to be ancillary to some other service.

-         The degree to which there are access links between the relevant audio visual material and other content. For example, audio visual material that is an ancillary part of an electronic newspaper is more likely to contain a significant number of links between that material and written articles in the nature of “click to read story/view video”, and/or of videos embedded in written articles.

-         The degree to which there are content links between audiovisual material and other content. For example, audio visual material that is an ancillary part of an electronic newspaper is more likely to be the basis or subject matter of written articles, to be an audiovisual version of the written article or an amplified or enhanced experience of that article; and the audio visual material does not need to be watched for the user to receive the information the service seeks to convey.

-         Where the service provides audiovisual and written material:

(a)  the balance between the two kinds of material, in terms of quantity and/or prominence

(b)  whether the written material is brief and/or merely an introduction to, or summary of, the audio visual material or, by contrast, has significant length and depth; and

(c)   whether the audiovisual material is the primary means of conveying to users the information sought to be conveyed

-         Whether on an overall assessment the audiovisual material can or cannot be said to be integrated into, or ancillary to, another service. In particular, whether the material can reasonably be described as intended specifically to allow the viewing public to benefit fully from, or to interact with, information provided as part of some other, primary, service.

None of the specific factors is determinative. The service has to be considered in the round.   For instance, even though the Sun Video section was at one time branded as Sun TV, that alone was not enough to bring it within regulatory scope; nor was the fact that the Sun Video section had a cataloguing function.

While the OFCOM ruling leaves open the possibility that in some factual circumstances a newspaper site might contain a separate regulated video service, it does significantly restrict ATVOD’s regulatory reach.  Following OFCOM’s ruling the ATVOD Chief Executive Pete Johnson said:

“Most people will recognise that defining the scope of new regulations in a fast-moving market is a complex and difficult task. The appeal system is a vital part of the process, giving users and providers of video on demand services greater clarity over where the new protections for consumers do and do not apply. … We will now reflect further on the appeal judgement and consider any implications it may have for any other past and future rulings on whether a service falls within ATVOD’s remit.”

The Ready Relevance Reckoner

You don’t know what predictive coding is?  Well that may not be too surprising, since one thing that almost all the contributors to an October 2010 Survey on Predictive Coding (published by the eDiscovery Institute) agreed on was that there should be a better term than predictive coding to describe what they do. 

That is a shame, since predictive coding technology promises to be the most important advance in tackling the mountain of electronic documents that we litigators now face when doing discovery. (Sceptical contrary view here)

In a nutshell, predictive coding tools enable the lawyers to review a sample of the overall document collection and decide whether each document in the sample is relevant or irrelevant.  The tool analyses the reviewers’ decisions and then evaluates the rest of the collection, giving every document in the collection a relevance ranking. 

Each tool has a slightly different approach, for instance some are more iterative, but that’s the gist.  The hope is that since the vast majority of any given electronic document collection is irrelevant, most of the documents can with a high degree of confidence be excluded from human review – an approach that chimes very nicely with the emphasis in the English procedure rules on proportionality.

So what can we do about the name?

My own flight of fancy is the title of this post.  But it is hardly a serious generic term, though you never know – relevance reckoners, anyone?.  But even that doesn't convey the message that these tools enable the user to identify irrelevance, rather than relevance.  Since the key point of the technology from the human point of view is the ability to start with a subset and end up discarding most of the documents, maybe a suitable generic term would flow from that – sample-driven culling, perhaps?  Irrelevance cullers? Garbage detectors?

Hmm, maybe we’re stuck with predictive coding after all.  It has to be better than one of the alternatives suggested by a contributor to the eDiscovery Institute Survey: ‘Prognostic Document Profiling’.

[Updated 1 May 2012.  Predictive Coding now seems to be morphing into Computer Assisted Review.  Not perfect, but much better.]

SABAM/Scarlet meets Newzbin2 – but will they play nicely together?

The ECJ’s decision in SABAM v Scarlet is a seminal judgment. 

It is not just what it says, but the way that it says it.  By the standards of ECJ judgments it is unusually forthright:  IP rights are not absolutely protected or inviolable, but have to be balanced against other fundamental rights - not just as a matter of form but in substance. It is not enough to pay lip service to rights of privacy, freedom of expression and (the new kid on the block) freedom to conduct a business. 

The ECJ has stated in the plainest terms that these rights have real substance and can trump overreaching IP rights.  Moreover, the ECJ has given real teeth to the prohibition on general monitoring obligations in Art 15 of the ECommerce Directive.

SABAM/Scarlet comes hot on the heels of Mr Justice Arnold’s second judgment in Newzbin2.  There are important factual differences between the cases.  SABAM/Scarlet is about content filtering, whereas Newzbin2 is about site blocking.  But there are also parallels.  Both concerned conduit ISPs.  Both involved the use of DPI (deep packet inspection) technology, albeit in very different ways.  Both required the court to balance the various fundamental rights at stake.  How does Mr Justice Arnold’s approach now stack up against that of the ECJ? 

General monitoring obligations

Article 15 of the ECommerce Directive prohibits national authorities (which includes courts)  from adopting measures that would require an ISP to carry out general monitoring of the information that it transmits on its network.  In L’Oreal v eBay the ECJ held that this applies to measures that would require an online intermediary, such as an ISP, to actively monitor all of the data of each of its customers in order to prevent any future infringement of intellectual property rights.

The filtering system at issue in SABAM/Scarlet would require an ISP:

1.       to identify, within all of the electronic communications of all its customers, files relating to peer-to-peer traffic;

2.      to identify, within that traffic, files containing works in which rightsholders claimed to hold rights;

3.      to determine which of those files were being shared unlawfully; and

4.      to block file sharing that it considered to be unlawful.

This involved a cascade of deeper and deeper dives into transmitted packets in order to determine their nature and contents.  In the words of the ECJ:

“Preventive monitoring of this kind would thus require active observation of all electronic communications conducted on the network of the ISP concerned and, consequently, would encompass all information to be transmitted and all customers using that network.”

How does Newzbin2 sit with this?  The system of site blocking that the court required BT to employ was a two stage process: first, redirection based on IP addresses, then the redirected traffic would be analysed at the URL level. 

Although this would use DPI technology, in his original judgment Arnold J had taken care to distinguish between three levels of packet inspection, which he described as (i) a minimal analysis; (ii) summary analysis (iii) detailed, invasive analysis of the contents of a data packet.  The order against BT requires it to use:

“DPI-based URL blocking utilising at least summary analysis in respect of each and every URL  … .

2. For the avoidance of doubt paragraph 1 … does not require the Respondent to adopt DPI-based URL blocking utilising detailed analysis.”

Arnold J commented:

“The order sought by the Studios does not require BT to engage in active monitoring of the kind described by the Court of Justice [in L’Oreal v eBay] at [139], but simply to block (or at least impede) access to the Newzbin2 website by automated means that do not involve detailed inspection of the data of any of BT's subscribers. To the extent that this amounts to monitoring, it is specific rather than general.

It is clear from SABAM/Scarlet that (a) the fact that the deepest inspection is carried out only on some preselected traffic does not prevent there being a violation of Article 15 and (b) the deeper the dive into the packets, the greater the likelihood that Article 15 will be contravened.

It is also evident that Arnold J was alive to these issues.  However because the Newzbin2 judgment gives no further detail as to the difference between minimal, summary and detailed DPI analysis it is unclear exactly what level of packet inspection is involved and whether SABAM/Scarlet would have had any impact on that.

Arnold J went on to say:

“Furthermore, it would be imposed by a case-specific order made under national legislation which implements Article 8(3) of the Information Society Directive.”

However that, in the light of SABAM/Scarlet, is not an answer to a potential violation of Article 15.

Fundamental rights

Much of the fundamental rights analysis in SABAM/Scarlet and Newzbin2 is similar.  Both balanced the rightsholders’ IP rights against freedom of expression.  Both recognised that the risk of preventing access to lawful content through overblocking or overfiltering is a relevant factor to take into account. 

However SABAM/Scarlet not only placed heavy emphasis on the fact that IP rights are neither absolutely protectable nor inviolable, but also introduced some new elements.

The biggest contrast with Newzbin2 is in the ECJ’s reliance on the fundamental freedom of ISPs to conduct their businesses.  This is now enshrined in Art 16 of the EU Charter of Fundamental Rights: “The freedom to conduct a business in accordance with Community law and national laws and practices is recognised.” 

The ECJ held that the injunction in issue would be a serious infringement of the freedom of the ISP concerned to conduct its business since it would require that ISP to install a complicated, costly, permanent computer system at its own expense.

On the facts this differed from Newzbin2, in which Arnold J held that the costs of implementing the injunction were modest.  But he also said this:

“The Studios are enforcing their legal and proprietary rights as copyright owners and exclusive licensees, and more specifically their right to relief under Article 8(3). BT is a commercial enterprise which makes a profit from the provision of the services which the operators and users of Newzbin2 use to infringe the Studios' copyright. As such, the costs of implementing the order can be regarded as a cost of carrying on that business.”

Even before SABAM/Scarlet, this passage looked a bit lopsided.  The Studios also are commercial enterprises which make a profit from exploiting their legal and proprietary rights as copyright owners and exclusive licensees.  As such the costs of protecting and enforcing those rights, including implementing this order, could just as easily be regarded as a cost of carrying on that business.  In any event now, after SABAM/Scarlet, it will not be possible to elevate rights owners above mere commercial enterprises.  They both have fundamental rights which have to be balanced.

One point which SABAM/Scarlet leaves open is how the balance between competing fundamental rights is to be drawn if the costs burden of complying with an injunction is shifted to the rightsholder.  It could be argued that even if the compliance burden could be fully compensated by cost-shifting, an excessive burden is damaging to the interests of the internal market and consumers generally, whichever party bears it.  In any event cost-shifting cannot reduce any impact on the freedom of expression rights of ISP customers, which the ECJ emphasised, or affect violation of Art 15 of the ECommerce Directive.

The ECJ also noted a cross-border issue which is potentially of great importance when considering whether an injunction could result in overblocking and thus interfere with the freedom of expression rights of the ISP’s customers.  The court said:

“it is not contested that the reply to the question whether a transmission is lawful also depends on the application of statutory exceptions to copyright which vary from one Member State to another. Moreover, in some Member States certain works fall within the public domain or can be posted online free of charge by the authors concerned.”

The discussion in Newzbin2 was entirely domestic, predicated on infringements of UK copyright.  It now appears that future cases will have to take cross-border issues into account, at least where an ISP has customers in more than one EU country.

How dare we speak across borders!

I've been inspired by this excellent but thoroughly depressing piece by Nate Anderson in Ars Technica to add the "through any media and regardless of frontiers" quote from Article 19 of the UDHR to my sidebar.  Adopted in 1948, it could not have been better expressed had it been written with the internet in mind.

Anderson highlights the proposed US Stop Online Piracy Act (SOPA) as the first legislation that would officially distinguish between domestic and foreign internet sites, and the last nail in the coffin of the borderless internet. 

It has become de rigeur over the last 10 years not only to dismiss the early cyberlibertarians (Johnson, Post, Perry Barlow et al) as impossibly naive and idealistic, and to predict the erection of national borders in cyberspace, but also to welcome that as a Good Thing (Goldsmith and Wu). 

There is certainly a serious debate to be had over how far attempts by national authorities to prevent undesirable bits and bytes from flying across their borders can be effective; and about the negative externalities associated with going to ever greater lengths in trying to make them so.  SOPA is a case in point. 

But it's the last proposition that really sticks in the throat - the idea that we should, as a laudable goal, positively seek to re-erect national borders in cyberspace rather than take the opportunity provided by the internet to break them down. 

For domestic politicians wedded to notions of national sovereignty the idea of preserving national borders online is inevitably meat and drink.  The unfortunate and surprising thing is that these ideas seem to have traction beyond that (I had a slightly crisp exchange with an adherent at the Society for Computers and Law Policy Forum in September).  Some seem quite happy, in the name of preserving domestic sovereignty, to toss on the scrapheap the most liberating development for cross-border freedom of expression that has ever taken place.

Naive and idealistic?  Maybe, but unless we cling tenaciously to the rock of Article 19 we risk being swept away in the rush to put speakers back in their rightful and respectful place behind national barriers.  How dare we speak across borders!  (How dare we speak at all?)

ECommerce Directive trumps Jurisdiction Regulation (maybe)

The CJEU’s Martinez/eDate judgment (C-509/09 and C-161/10, 25 October 2011) is another significant decision on cross-border internet liability.  In December last year we had Pammer/Alpenhof, which provided much needed guidance on what constitutes cross-border targeting of online activities under the consumer contract provisions of the Brussels Jurisdiction Regulation. 

Now the Court has turned its attention to cross-border tort liability, in two joined cases that raise questions under both the Jurisdiction Regulation and the Electronic Commerce Directive.   

The eDate case involved a claim for violation of personality right, the Martinez case infringement of privacy.  eDate was an Austrian company sued in Germany over statements on its Austrian website.  Martinez sued the English publisher MGN in France, over a publication on its English Sunday Mirror website.   

The most heavily reported aspect of the judgment has been that on jurisdiction.  I will start at the other end, with an aspect that didn’t even make it into the Court’s press release – and yet which has equally far-reaching implications.  This concerns the scope of the country of origin and internal market clauses of the ECommerce Directive.  These are powerful, often overlooked, provisions which can require a court to disapply a Member State’s otherwise applicable local law.

Some have argued that it is at best unclear whether these clauses apply only to restrictions of a public law or regulatory nature, or can also include disputes over private rights.  The judgment confirms that these clauses do apply to private rights. 

The country of origin provision of the Directive (Art 3.1) requires a Member State to ensure that service providers established in its territory comply with its law.  The internal market clause (Art 3.2) prohibits a Member State from restricting the freedom to provide information society services from another Member State.   Both apply only within the ‘co-ordinated field’ of the Directive, and are subject to exceptions such as copyright and industrial property rights. 

One argument in favour of limiting these provisions to regulatory regimes is the reference in the definition of co-ordinated field to ‘requirements with which the service provider has to comply’ – a phrase that does not read easily onto private law rights and obligations.   

The CJEU devotes two paragraphs of its judgment to determining that both the law to which a service provider established in a Member State is subject, and the provisions requiring other Member States to respect the binding nature of that law, include the private law field.  The Court relies on the reference to private law disputes in Recital (25), the reference to excluded private law rights and the reference to liability of service providers in the definition of the co-ordinated field.  The result is that the provisions apply to liabilities such as defamation and privacy infringement.

This should be very significant.  It ought to provide a basis on which to prevent a stringent local law from applying to an online service incoming from another Member State with a more liberal law.

That is exactly what the CJEU went on to hold.  The country of origin and internal market provisions do not amount to a rule of conflicts of law.  But the Member State must ensure that within the coordinated field, and subject to certain permitted derogations in a specific case, the provider of an electronic commerce service is not made subject to stricter requirements than those provided for by the substantive law applicable in the Member State in which that service provider is established.

This is all very positive for cross-border trade and publication on the internet within Europe.  But the last aspect of the judgment, the one that garnered all the headlines, is not.  It allows a person claiming infringement of personality rights to sue under Article 5(3) of the Brussels Regulation in the Member State in which his centre of interests is based (as an alternative to the Member State of the establishment of the publisher), for all the damage caused; or to sue in any Member State where the online content is accessible, only for the damage in that country.  So it is now very easy for a publisher established in one Member State to be sued in another, for the contents of a merely accessible website.  That is bad news for the internet.

It is still tempting to ask “so what?”.   According to the remainder of the judgment, under the ECommerce Directive the court should consider whether its local law is more stringent than that of another Member State in which the defendant publisher is established and, if so, disapply its local law unless grounds for a specific derogation can be established.  That is the theory.  In practice it is difficult to avoid the suspicion that a defendant would be better off arguing in front of its home court, rather than trying to persuade a foreign court to disapply its own local law.