Guest Post Volker Rieck

A year later: A review of the campaign against EU copyright reform

More recent findings allow a closer look at the campaigns, their backgrounds and, in particular, the key players in financing them.

About a year ago, the German protests against the reform of EU copyright law began. A group of about 120 demonstrators gathered in front of the Brandenburg Gate, but they had some bad luck with the weather; June 22 was one of the few rainy days in the otherwise very dry German summer of 2018.

As reported in this blog, the discrepancy between the size of these and other demonstrations and the sheer mass of mails and tweets received by MEPs in July and August 2018 gave rise to considerable doubts about the actual size of the alleged protest movement. The fact that the protests themselves had a real core was not questioned by anyone – not even by reports published here. However much more was made of the online tools used at the time, because there were, for instance, no verifications. These tools could easily be controlled by automated scripts as we demonstrated several times on this website.  And indeed, there were many indications that many of the tweets and emails were not organic.  For instance, why did number of tweets remain steady or increased in overnight hours in Europe.  Why were tweets directed a specific country’s MEPs and then suddenly change to another set of MEPs?

Nevertheless, such facts seem to have little relevance for Green politicians* like Tabea Rössner – in her statement made in a debate in the Bundestag on 13.03.2019 she stated that we claimed in our reports “bots” were responsible for the protests and couldn’t be explained otherwise.

This is not only factually wrong, but also foolish. No such claim was ever made.

While Ms Rössner puts forward these straw man arguments she also successfully represses other facts.  For example, she fails to note that she herself gave a speech at the “Day of Action” against the Directive at the end of August 2018 in Mainz in front of just 10 demonstrators, while MEPs were flooded with emails and tweets.

2017: Origins of the campaigns

In order to better understand the campaign as a whole, it is worth taking a somewhat broader look back. In spring 2017 the terms Censorship-Machines and Linktax appeared at the Canadian NGO Open Media in connection with the EU directive.

Illustration: OpenMedia promoted the slogans “link tax” and “censorship machines” in March 2017 including a webform to contact MEPs.

Unfortunately, Open Media is anything but transparent with regard to its financing. The last audited annual report dates back to 2016. Platinum sponsors like Mozilla mysteriously disappear from the website, and financial reports were temporarily removed from the site. Who really financed Open Media in 2017 and beyond remains unclear. One should note Open Media’s board does includes Jacob Glick formerly Public Policy chief for Google in Canada.

Shortly after Open Media introduced the framing of the terms Censorship-Machines and Linktax for their campaign in the EU, the Internet conference Re:Publica took place in Belin, where a panel on Censorship-Machines was shown.

Illustration: Youtube video of the panel from 17.05.2017

The composition of the panel is particularly interesting because 3 of the 4 panel participants or their employers should play a decisive role in the later course of the campaign.

The participants were present:

– Jake Beaumont-Nesbitt, International Music Manager Forum, rather an exotic in the group, but his association is a member of Copyright for Creativity (C4C).

– Raegan MacDonald, Senior EU Policy Manager of Mozilla,

– Diego Naranjo, Senior Policy Advisor at EDRi

– Caroline de Cock, who is described on the Re:Publica websites as coordinator of Copyright for Creativity (C4C); that she is also managing director of the Belgian lobby company N-Square, was not mentioned at all.

Illustration: Re:Publica presentation pages of the panel participants

Mozilla was demonstrably one of the financiers of Open Media, which in turn, through its subsidiary New/Mode, provided the “engagement tools” with which MEPs were flooded with e-mails and tweets in the summer of 2018.

Mozilla sponsored financially and technically the telephones calls to MEPs. The webform that auto connected callers helpfully included a script to read to MEPs and their representatives.

In autumn 2018, however, Mozilla suddenly disappeared from the Open Media sponsor list.

As was already known, it was Caroline de Cock who registered the website saveyourinternet.eu in May 2018. At no time did the site itself have a legally valid legal notice in accordance with the E-Commerce Directive, but merely referred to C4C. At some point at the end of 2018 C4C disappeared from the site and was replaced by the note “managed by EDRi”, which also does not fulfill the legal notice obligation. The site also began to use the anonymization service Gandi. It seems obvious that the actual owner or operator of the site was persuaded by someone to better conceal the ownership of the site.

Moreover, saveyourinternet.eu completely lacks data protection notices, which are also mandatory under the Basic Data Protection Regulation.

The lynchpin of the relevant campaigns: a US industry association!

The “ad hoc coalition” C4C (own statement) with the “coordinator” de Cock counts the US industry association CCIA (Computer and Communication Industry Association) among its donors.

However, the CCIA also financed the campaign site Create.Refresh, which had demonstrably targeted YouTubers since the summer of 2018 in order to have videos created against the (what was then called) Article 13 of the Copyright Directive. In practice, this procedure included talking points which, although wrong, were bluntly adopted by several prominent YouTubers.
Research from the German newspaper FAZ also found that some YouTubers were offered money to create such videos.

Create.Refresh may also have invested a large part of its money in the campaign via Twitter. From August/September 2018 on, it was impossible to avoid the sponsored Twitter posts anymore, especially since the German Pirate Party MEP Julia Reda happily retweeted these sponsored tweets – which, I might add,  are political advertising.

Illustration: Retweet of a campaign contribution of Create.Refreh by Julia Reda in August 2018

The CCIA can be described as a key player in the campaign. They not only used to finance C4C, saveyourinternet.eu and Create.Refresh, but also to provide substantial parts of the N-Square/KDC Group 2018 budget, as the EU Transparency Report stated at the end of June 2019.

Illustration: Extract from the EU Transparency Register, N-Square/KDC Group Financing 2018

As payments in the EU Transparency Report only have to be reported in funding level brackets,the CCIA’s share of N-Square’s business may well have been significant.

 

No less than four organizations that played a decisive role in campaigns against the copyright reform received money from a US industry association that tried to exert a massive influence on EU legislation and at the same time to sell these efforts as civil society resistance (grassroots movement).

 

Friends of science

To complete the picture, at the Re:Publica 2019 Prof. Kretschmer from the University of Glasgow gave a lecture on EU copyright reform.

Illustration: Prof. Martin Kretschmer at the Re:Publica 2019, source YouTube

Prof. Kretschmer expressed doubts about the evidence presented by myself, this blog, The Times of London and other newspapers on the funding and actions of the groups identified in this article.  This despite the fact all the evidence comes from the groups themselves or is derived from the EU transparency database. And the actions of these groups are proudly proclaimed by the groups themselves.  In the American vernacular Kretschmer acts as a “merchant of doubt” despite the fact there seems to be no doubt to sell.

Almost inevitably, however, the question arises about Prof. Kretschmer’s cooperation with N-Square, which as noted here was of central importance in the campaign. Did it cease in 2017?

 

Illustration: Website of the University of Glasgow about Prof. Martin Kretschmer and his projects in June 2017

It can therefore only be speculated whether his statement, heard in the video, that he was “not completely innocent” in the debate was possibly meant quite differently.

Prof. Kretschmer only knows the truth himself.

 

Editors Note: In light of the very recent article by Frankfurter Allgemeine Zeitung on efforts by cyberturfers Create.Refresh to pay YouTubers to influence debate on the EU Copyright Directive, this analysis of influencers on Twitter is extraordinarily important.  Please note, this article was reworked from earlier version, omitting location data that proved to be unreliable).

Who tweets on EU copyright?

The debate on the reform of EU copyright law has many facets and a considerable part of the discussion takes place on Twitter.
The power of hashtags

On the one hand, hashtags are used to search for key words in discussions particularly quickly and effectively. On the other hand, it is very easy to find out which topics receive special attention on Twitter by analyzing which terms are often hashed and how often and dynamically they are shared. These terms are then displayed as so-called “trending topics” on the Twitter home page. The so-called “retweeting”, i.e. the unchanged forwarding of a tweet along with hashtags to one’s own followers, is an important factor here.

The frequency of use and distribution of hashtags is therefore often seen as an indicator of the public relevance of certain topics. If, for example, the #saveyourinternet hash tag appears regularly in the trending topics, the impression is created that more and more people are rejecting the copyright reform.

The importance of influencers

Now that the final decision on the reform is about to be taken, it makes sense to take a closer look at the Twitter campaign against the EU copyright reform using aggregated data. The data is based on data from the Talkwalker service, which we have compared with alternative services such as Keyhole. Here we use only the global data without geolocation, i.e. the worldwide number of uses of certain hashtags in a certain period of time, were examined. (the geolocation data proved to be faulty).

Using data from the Hashtagify service, the largest influencers were measured in terms of range.
This makes it easy to see who gave a hash tag the most momentum.The so-called “mentions”, i.e. the mention of a hash tag in a tweet, are decisive for the spread of a hash tag. The more mentions a hashtag experiences, the more it is perceived as relevant for a topic.

The following analysis focuses on the mentions of the hashtags, which play a major role in the campaign against the EU directive from a German perspective, because they are the framing of the anti-reform movement.

#saveyourinternet

#Artikel 13

#upload filter

#linktax

#censorshipmachines

The temporal development of mentions

Most noticeable is the increase in mentions of #saveyourinternet after Google/Youtube started its active campaign around this hashtag in autumn 2018.  The numbers doubled.

Illustration: worldwide cumulative representation of the development of the hashtag #saveyourinternet. Source: Talkwalker

The  graph illustrates very well the influence that Google has on the debate.

The timing  of the votes in the parliament created short peaks peaks in the change (or delta) in number mentions of  #saveyourinternetcampaign.  These boosts helped prolong and expand the reach of the campaign.

Illustration: Overall development of the hashtag #saveyourinternet 01.2018- 02.2019 on a weekly basis, source: Talkwalker

One can see very well in this weekly development the impact that Google had through its own Twitter channels Youtube and Youtube Creators.

The campaign with the Hashtag #saveyourinternetwas mainly determined by Google and Youtube from the end of October and was much broader in November/December than in the summer of 2018. There were, however, also peaks that could be explained by the two voting dates.

This is confirmed by the reach measured by the Hashtagify service. It is referred to there as Influence, which we refer to in the following charts and in the text as reach.

In this analysis, the top 10 influencers for different hashtags were examined, but this also means that there were of course far more influencers involved.

The top 10 influencers at Hashtag #saveyourinternettherefore had a reach of 1.5 billion

However Youtube and YoutubeCreators constitute 1,3 billion at the reach.

That corresponds to 85% of the Top 10 Influencer.

7 further of the Top 10 Influencer are Youtuber. It should be noted that Twitter accounts from Youtubers sometimes have a slightly different name than the Youtube channels. There are also Youtubers who have several Twitter accounts.

The only outlier is the Pirate Party, with a very small percentage.

The second largest reach in this study was hashtag #Artikel13.

However, it was only 9% of the range of #saveyourinternetwith 135 million reach (Top 10 Influencer number)

6 of the top 10 influencers are Youtuber. Other media also play a role here, as the shares of the Heuteshow, Extra3 and the SZ illustrate.

With a reach of 111 million for the top 10 influencers according to Hashtagify, the hashtag #Uploadfilter comes in 3rd in this review. 5 of the Top 10 are Youtuber.
Other media like Heuteshow, ZDFheute and SZ are also among the Top 10.

The Hashtag #linktaxcomes to a reach of 24 million.

So it seems to have captured less attention.
There is a somewhat different picture.
The US organizations Creative Commons and EFF apparently managed to best spread this hashtag in the debate about an EU legislation.

The Hashtag #censorshipmachinesachieved the lowest reach with 8.4 million related to the Top 10. Nevertheless it is remarkable that a US organization (Creative Commons) with almost 70% share within the top 10 dominates the debate in a European legislative process with an anti-reform hashtag.

Conclusion:

There are two inescapable lessons:
1.  fear sells
2.  YouTube and Google can effectively use their platforms to capture the attention of their users and to directly affect perceptions on matters relevant to their operations. This capacity can be used to distort debates and create an appearance of movements, or can actually create movements based on the presentation of one-sided information that distorts what is actually a far more nuanced truth.

YouTube intentionally created panic about its future that rather predictably resulted in an outpouring of expression of concern, reflected in the data presented here. A cynical campaign that itself illustrates the dangers to society presented by monopoly/monopsony platforms.

 

Translated from German.  Original here.

Cloudflare: The bad, the worse and the criminal

In the US, a large technology company is about to go public. Cloudflare, a San Francisco-based company, wants to collect nearly $3.5 billion on the stock exchange in the first half of the year with the support of the investment bank Goldman Sachs. But there are dark shadows over Cloudflare. The spectrum of its customers ranges from credit card fraudsters and spammers, to sites that engage in copyright infringement as a business model, to terrorist sites. Even US embargos are undermined.

What is Cloudflare?

Cloudflare offers a content delivery network.

In simple terms, it provides a kind of turbo drive for web pages, allowing them to be delivered world-wide quickly and securely. Cloudflare places itself between, on one hand, the web page and/or servers of its customers and, on the other, the site visitor and/or user of the service. By enabling it to selectively control and distribute site traffic, it can offer increased speed and protection against network overload attacks (DDoS).

However, Cloudflare also offers another feature: anonymizing its customers.

By placing a virtual screen over the original web page and/or their server, Cloudflare makes the operator practically untraceable. Upon inquiry, Cloudflare will only provide its own data, hiding client information such as hosting service and IP address, making it impossible to take legal action against illicit sites and services.

Civil law inquiries are futile, because Cloudflare only provides the naming of the hosting services, which is worthless without the respective IP address. This is roughly equivalent to seeking info on an unmarked apartment with just the address of a high-rise building housing thousands of residents.

The Cloudflare problem is well known

This anonymizing feature from Cloudflare attracts a number of unsavory customers including, again and again, copyright infringers.  But it doesn’t stop there.
Since December 2018, the EU Commission has included Cloudflare on a watch list for counterfeiting and piracy. Most recently, the service received the dubious prize as the worst enemy of the creative community from the US blog TheTrichordist.

The listing of infringing market participants has a long history in the US.
The music association RIAA submits an annual list of the worst offenders to the US Department of Commerce. In 2017, 9 out of 20 violators could not be identified by the RIAA because Cloudflare effectively camouflaged them.
The US film association MPAA is also aware of the problems with Cloudflare obfuscation and names the company in its annual list of interferers.

In the relatively new piracy segment IPTV – the streaming of non-licensed TV signals – the company is also on the move. A study from Fall 2018 shows the role of Cloudflare both in camouflaging the sites that sell IPTV subscriptions and in concealing the origin of the streams.

In a survey of data centers comprising file and streaming hosts in 2016, 40% of the Top 10 and 47% of the Top 30 used Cloudflare.
The ECO, a German association, which obviously doesn’t care about anything

 

Cloudflare is a member of the German industry association ECO. The purpose of this membership is probably to get a discount for traffic at the Frankfurt (DE-CIX) internet node, which ECO operates through a subsidiary.
ECO has never seemed to care that providers who are very heavily involved in piracy, including Cloudflare, are members of the association. In any case, there was no reaction to corresponding reports that ECO members, including Cloudflare, are responsible for over 50% of piracy traffic in the film sector in 2014, with 45.2% of this activity accounted for by Cloudflare and around 6% by a further five members.

Screenshot: Extract from the ECO member list, February 2018,

www.eco.de/ueber-eco/mitglieder/#C

Cloudflare in court

The reports of legal proceedings against Cloudflare are long and concern more than just virtual goods. For example, two manufacturers of bridal fashions filed suit for trademark and copyright infringements by plagiarizers who were made anonymous by Cloudflare. And, while a claim brought by adult entertainment provider ALS-Scan ultimately ended in settlement, the judge found that Cloudflare’s activities could significantly support copyright infringement by hosting cached copies of files (though the settlement precluded a final judgment on Cloudflare’s actions and liability).

Supporting Illegal Activity: Calculated or Coincidence?

In Fall 2018, Cloudflare made news by ending its business relationship with pirate hosts like Rapidvideo due to violating its terms of use. Such action begs the question: is the service being cleaned up in anticipation of the IPO or are these just insignificant and cynical platitudes? After all, before this,  Cloudflare had only voluntarily terminated its business relationship with US Nazi site the Daily Stormer in 2017.

Screenshot Youtube Video with Cloudflare CEO Matthew Prince on Fox Business Network

Big Data brings it to light

The current Google Transparency Report offers a look at the actual extent of Cloudflare’s involvement in piracy.
In the report, Google lists all requests from rights holders for deletions from the Google search index that concern rights violations.

In order to understand the importance of Cloudflare to the market, the 5,000 domains that are inactive sites parked with companies such as Team Internet, Sedo or GoDaddy but still protected by Cloudflare have to be subtracted from the 5,000 domains.

This leaves 3,645 domains. Of these 3,645 right-infringing sites, 41.9 % run via Cloudflare. For their part, they are responsible for 44.7 % of the copyright infringements reported to Google.

If one were to extrapolate this proportion to the total number of domains listed in the Google Report for copyright infringements, one would come up with almost 670,000 domains protected by Cloudflare – a significant portion of the 2.2 million domains with requests for delisting from Google’s search engine.

 

Among Cloudflare’s customers are:

Torrentz.eu, Gosong.net, Share-online.biz, Catshare.net, Bitnoop.com, Deepwarez.org, Turbobit.net, Myfreemp3.eu, and Nitroflare.com.

Each of these websites received at least 3 million deletion requests from the Google search index.

Not only pirates love Cloudflare – also credit card fraudsters, phishing sites, extortionists, and terrorists

The Watchwebsite Crimeflare is a real treasure trove of information about Cloudflare, liisting650 credit card fraud sides alone, to which Cloudflare offers shelter.

 

Cloudflare also proudly deals in SSL certificates, providing sites like Phishingseiten the manufactured consumer security and confidence-building necessary to be successful. According to the German magazine Heise, hundreds of such certificates for cheats were issued by Cloudflare.

Of course, as Spamhaus reports, the spreading of Malware often takes place over Cloudflare.

With Cloudflare, extortion is also par for the course, which conveniently generates additional services. By providing anonymity and untraceability to sites threatening, for example, to bring a web page to a standstill through DDoS, Cloudflare can then sell the attacked site its protection services.
A truly special form of customer acquisition.

Cloudflare has also found good business in terror.

As far back as 2012, the news agency Reuters confronted Cloudflare with the fact that it maintained the websites of Hamas and Al-Quds, designated by the US as terrorist groups.
In 2015, a petition against Cloudflare in the US highlighted that the service offered shelter to about 50 websites attributed to ISIS.
And in 2018, terrorist organizations were still being supported, with Dutch security researcher Bert Hubert identifying at least 7 different terrorist organization websites that use Cloudflare.

The Huffingtonpost had these findings evaluated by Benjamin Wittes, Senior Fellow of the Brookings Institution:

“This is not a content-based issue. Cloudflare] can be as pure-free-speech people as they want – they have an arguable position that it’s not their job to decide what speech is worthy and what speech is not – but there is a law, a criminal statute, that says that you are not allowed to give services to designated foreign terrorist organizations. Full stop.”

As icing on the cake, the company even has customers who are on the official embargo list of the US (SDN list). For example, CENTRAL REPUBLIC BANK from the Donetsk region uses Cloudflare’s services.

Screenshot: Collage of information from theUS Treasury’s Office of Foreign Assets Control
https://www.treasury.gov/resource-center/sanctions/SDN-List/Pages/default.aspx

Screenshot: Whois record of crb.dnr.ru at February 18^th2019

Do investors actually know what they are investing in?

Against the background of all these facts, two things are worth considering:

1) How has Cloudflare been able to obtain financing rounds from various investment companies in the past, including Google’s parent Alphabet?

2) Does Goldman Sachs actually know anything about the extent of its involvement in rights violations and its support of dubious “ventures,” even to the point of undermining US embargoes?

Risk management is one of the central parameters of investment banks when evaluating investments. Risks must be known and assessable in advance. Cloudflare’s considerable participation in dubious transactions is rare in an IPO and a huge risk. Particularly if, as in the ALS-Scan case, the company is faced with its own liability, or if criminal law is violated through the service’s business with terrorist organizations.

Goldman Sachs and current investors either lack moral standards, are naïve, or consider the risk of failure to be very low, which only shows how urgently we need government regulation of intermediaries on the Internet.