EU wants tech firms to police internet

Internet platforms will have to weigh laws in each member state, understand and apply their respective national court cases on hate speech, and then decide if the content can be removed.

"They have to understand the language and they have to understand what is prohibited by the national law and what is defined as hate speech in the case law in the decisions of the courts in each member state," said Vera Jourova, the EU commissioner for justice.

She noted that the Strasbourg-based European Court of Human Rights clarifies free speech to include the right to offend, to shock, to disturb the state or any part of the population.

But the announcement also follows Spain's blockade and removal of pro-Catalan websites in the lead up to an independence referendum, planned for 1 October.

Andrus Ansip, the EU commission vice-president for the digital single market, told reporters that online freedom of expression must be safeguarded, but then only gave short shrift to complaints about Spain's website assault.

"Talking about Spain - I fully respect the territorial integrity of the Kingdom of Spain, I fully respect the sovereignty of the Kingdom of Spain," he said.

European Digital Rights (EDRi), a Brussels-based association of digital civil rights organisations, noted that in its draft communication, the EU commission had quickly removed references to domain name services having "a role to play in defining solutions" following the recent Spanish move.

"This is a very good example of how the simplistic 'solutions' proposed by the commission run into trouble as soon as they are faced with real-world situations," it said in a statement.

The Napalm girl and Facebook

While the commission's recommendation is broadly geared towards hate speech and terrorism, governments and internet firms in the EU have also been known to seize websites or content for reasons that appear entirely arbitrary.

Last year, Orange, a French telecoms firm, shut down Google and Wikipedia for an entire morning in France after adding them both to a terror watch list.

Facebook also drew controversy when it removed a historic photograph of a naked nine-year-old girl in Vietnam running away from a napalm attack. Facebook later reversed its decision.

YouTube once even removed a video debate between liberal Dutch MEP Marietje Schaake and EU trade commissioner Cecila Malmstroem. YouTube said the video had violated its community guidelines, without offering an explicit reason.

The EU commission wants the take downs to be automated, but this poses wider questions on legitimate content swept up in the purge against hate speech and terrorist propaganda throughout the differing legal jurisdictions among EU states.

It says a code of conduct, transparency, trusted flaggers, and online notice systems would help prevent abuse.

"Those who give the notifications should be informed of what happened - this is in line with our new transparency push. And, in relevant cases, even the authors of the comments should be informed that it was removed and why," noted Jourova.

But not everyone is convinced.

MEP Schaake, in an emailed statement, said it is extremely dangerous to rely on big tech companies to take automated measures in removing content.

"The commission should be pushing back against the trend, not embracing it. There can be no room for upload-filters or ex-ante censorship in the EU," she said.

Similar comments were made by German Green MEP Jan Philipp Albrecht, who said the commission's plan "gives room for those companies to make their own principles and definitions."

Earlier this week, Nils Muiznieks, the Council of Europe's human rights commissioner, had also warned against allowing states to rely on "private actors" to regulate online communications.

The commission is not opposed to legislating on the issue, but has deferred any decision to write up binding rules until next May.

Instead, it first wants to see how well tech companies manage the self-regulatory screenings - despite receiving requests from CEOs to regulate.

Jourova went to Silicon Valley in the US last week, and discussed the issue with some of the biggest tech firms - including Facebook and Google.

She said that the vast majority had no interest in policing the web and instead wanted clear regulations.

Many of the representatives of tech companies in Silicon Valley said that they "do not feel comfortable" being the ones who decide this, Jourova said.

She also did not rule out applying a German law across the EU, which imposes up to a €50-million fine on social media firms for failing to remove illegal hate speech and other posts within 24 hours of receiving a notification.

"We will look at this German case [and] how successful this law is," she told reporters earlier this week.