10 everyday things on the web the EU Commission wants to make illegal: Oettinger’s legacy

by Julia Reda

In a few days, scandal-prone Günther Oettinger will stop being Europe’s top internet policy maker – he’s being promoted to oversee the EU budget.

But before leaving, the outgoing Digital Commissioner submitted dangerous plans that undermine two core foundations of the internet: Links and file uploads. While Oettinger is going away, his lobby-dictated proposals are here to stay.

These proposals are pandering to the demands of some news publishers to charge search engines and social networks for sending traffic their way (yes, you read that right), as well as the music industry’s wish to be propped up in its negotiations with YouTube.

These proposals will cause major collateral damage – making many everyday habits on the web and many services you regularly use downright illegal, subject to fees or, at the very least, mired in legal uncertainty.

We can still stop these outrageous plans – but only if you demand from your representatives in the European Parliament that they join me in rejecting the proposals.

Here’s what may otherwise become illegal:

01 Sharing what happened 20 years ago

Sharing snippets of news articles e.g. on a blog or a personal website without a license from the publisher will be an infringement – even 20 years after the article was published.

The EU Commission has not proposed any exceptions for even the shortest of snippets, or for individuals, or for non-commercial purposes. It also doesn’t matter whether a link to the source is provided or not.

02 Tweeting a creative news headline

“Wir sind Papst” (We are Pope) is a famous headline by German tabloid Bild. Unless the person tweeting it pays Bild’s publisher Axel Springer for a license, tweeting this three-word headline would be an infringement of the proposed extra copyright for publishers.

Twitter could also pick up the tab, maybe paying a collecting agency for a blanket license, thereby freeing you from the duty to negotiate one – either way, an arrangement would have to be made.

03 Posting a blog post to social media

The preview pic and text snippet that Facebook, Twitter, Reddit and other services automatically generate when you share a link would be subject to licensing if that link goes to a “press publication” – which explicitly includes regularly updated entertainment websites.

If Facebook and Twitter don’t want to start paying for links, they may have to disable this function, making their interfaces less user-friendly.

04 Pinning a photo to an online shopping list

Visual bookmarking services like Pinterest allow you to grab images off web pages and save them in an organized way, for example to make a shopping list or just to gather inspiration.

In doing so, they copy and republish the title, image and a snippet of text from the page you found the image on – which will be an infringement of the extra copyright for publishers, if the EU Commission has its way.

05 A search engine indexing the web for you

To allow users to search the web, a search engine first has to “read” all websites using a robot and create a database of which content is found where. To be useful, such a database needs to include copies of copyrighted material.

The proposed law threatens to outlaw making such copies of press publications without licenses from the publishers. Just storing a copy of the data will be enough to require a license – it doesn’t matter whether anyone can read it, e.g. whether the search engine displays snippets of this stored text to users in their search results or not.

Bing, Google, Seznam.cz and others would have to get licenses from all journalistic sites – or a blanket license through a collecting society that might be set up for that purpose – or they would have to de-list them from their search results.

Since news websites get a lot of traffic from Google, they would probably move quickly to give Google a free licence to continue listing them, but smaller search engines may not be so lucky.

06 A portfolio hosting site not monitoring your uploads

FotoCommunity is a Germany-based social network for photographers, hosting millions of images uploaded by their rightholders.

Today, when someone reports that an image uploaded by one of FotoCommunity’s users infringes their copyright, FotoCommunity is obligated to take it down. In return, they are not liable for the infringement themselves.

The proposed reform would turn this on its head: FotoCommunity would now be obligated to actively prevent users from uploading anything rightholders had identified to FotoCommunity.

This is a huge engineering challenge: The site would need to develop a filter that checks each uploaded picture against a database of copyrighted pictures. YouTube, which runs a similar filter on uploaded videos to identify copyrighted music or clips, says it has spent more than 60 million dollars building this technology.

Even worse, FotoCommunity would be required to detect any kind of copyright infringement in the uploaded pictures. The simplest case would be that an identical copy of a protected picture were uploaded. But a picture can also infringe copyright of a sculptor or architect, if it’s a photograph of their work. Detecting a picture of a 3D artwork from any angle is much more difficult than matching uploads against a database of protected pictures — if not impossible.

07 Github allowing unmonitored commits

The obligation to scan all uploads for copyright infringement would apply to any kind of service that hosts “large amounts of works”, not just photos.

Since the EU Commission has not foreseen any exceptions, popular services that are not at all associated with widespread copyright infringement, such as software repository hosting service GitHub would nevertheless have to put in place the filtering technology to address a non-existent problem – at the latest, as soon as the first rightholder of source code identifies some code to them that they want to keep off the site.

European startups like MuseScore, which allows users to upload sheet music, might also need to develop technologies for detecting copyrighted instances of, in this case, sheet music or protected melodies. Such an obligation is likely to endanger their existence.

8 Wikipedia ACCEPTING unmonitored uploads

The obligation to scan all uploads would not just apply to commercial sites, but also to projects like Wikipedia that aren’t run for profit, and which expressly only allow uploads of photos licensed for public re-use. If found to be “providing access to large amounts of works uploaded by their users” they would still need to “prevent the availability of works identified by rightholders”.

On Wikipedia, volunteers may review newly uploaded images – but it’s doubtful whether this loose process would satisfy the new law. More likely, they would need to implement “effective content recognition technologies”.

09 Training your own artificial intelligence

Okay, you might not be doing this today yet… but someday soon you might want to.

As artificial intelligence becomes more pervasive, the way in which users make their computers fulfil useful tasks is changing: Traditionally, you program the computer to do a task by programming it – writing step-by-step instructions. An AI, however, is not programmed, but trained. You train it by giving it a lot of input representing what you want it to do for you, and letting it work out the required steps itself. Training an AI is likely to involve copying data – much of it under copyright.

The proposed copyright reform for the first time introduces a Europe-wide copyright exception for text and data mining, e.g. analysing a large amount of data – but only for “research institutions” and “for the purposes of scientific research”.

That doesn’t include you – or the countless other hobbyists, hackers, coders and amateur researchers who could be making valuable contributions and discoveries… or just using technology to learn and play.

Not affected: MegaUpload

Despite all the new restrictions on hyperlinks and uploads, sites like MegaUpload, which was famously shut down by US authorities for allegedly systematically infringing copyright, would not be affected.

That’s proof: This law is not aimed at sites that actually play fast and loose with copyright – it’s meant to get social networks and search engines to fork over money to struggling European cultural industries.

Look for details here

Leave a Reply

Your email address will not be published. Required fields are marked *