10 reality-challenged ways that the EU's departing internet commissioner tried to destroy the internet

Since 2014, we've chronicled the reality-challenged internet proposals of the scandal-haunted EU Commissioner Günther Oettinger; now, on the eve of Oettinger's promotion to EU budget chief, MEP Julia Reda lists the 10 normal web activities that Oettinger has sought to ban, from sharing snippets of 20 year old news articles to quoting three-word newspaper headlines to creating and operating a search engine.

Oettinger's career has been dominated by poorly thought-through, evidence-free plans to save EU newspapers by imposing a tax on linking to them, which would require banning all linking without payment, and would shut down any website that allowed its users to post material without first having a copyright expert review that material to make sure it wasn't an infringement of the special rights for newspapers — so say goodbye to Wikipedia and Github.


Oettinger's proposals lack any of the basic futureproofing you'd want to see from a law — for example, they would ban feeding archives of newspaper material to AIs to train their algorithms, a natural extension of the kinds of machine-learning techniques that have been deployed since the first days of the internet (for example, in spam filters).

Oettinger is being Peter-Principled out of the copyright/internet file, but his proposals live on, and are about to be voted on by the EU member states. Without Oettinger there to defend them, they are weak and vulnerable, and EU citizens can help give them the anonymous death they deserve.

7. 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".

10 everyday things on the web the EU Commission wants to make illegal: Oettinger's legacy
[Julia Reda]