How a law to reign in large platforms will end up costing large platforms least of all.
Policy making is, by nature, one step behind technology because it tends to focus on (and is lobbied by) today’s companies. When lawmakers, however, by virtue of universally applicable and EU-wide laws, try to come after a hand full of big players, society and smaller startups suffer. With the Copyright Directive, the EU risks shooting itself in the foot. Worse even, it is launching a torpedo at its own vision of becoming a startup continent. Here is the gist of it:
The proposal to filter online content fundamentally misses its aim. By targeting a few big video platforms, it will ultimately uplift and fence their market share. The scope of the proposal is flawed. While intending to govern only licensed content, it targets all types of content and all platforms regardless of licenses or copyright. Filtering itself is technically ineffective and will cause more damage than good on the internet. Seemingly easy on text through hashing but disproportionately expensive for anything more complex or even impossible. The suggested filtering technology will raise the cost of launching a startup in Europe and drive talent away. First-buried-then-leaked evidence suggest it will not solve the problem anyway. For all of us it will result in lower quality, less variety and content online, as the law favours those who delete content.
Let’s take this in turns. To start, the proposal aims to address a grievance by rightsholders, namely the fact that certain large platforms don’t pay as much as the content industry is wishing for. While it’s one thing whether we want laws in favour of individual industries, this proposal will actually not do anything to abate this discord. Ironically, this proposal makes it even more likely that small platforms and innovative startups pick up the tab, as they don’t have the market power and legal teams to go through thousands of licensing agreements. To be clear: Startup founders fully respect creation and its remuneration. But this law is drafted in entire disregard of Europe’s startups and its citizens’ fundamental rights.
Besides fundamentally missing the target, the proposal is carpet bombing the entire digital world. Regardless of whether one uses licensed content or not, everyone will now have to enact a costly regulatory prescription. Content can range from images over text, audio visual content, objects to code. While policy makers probably would have liked the idea of squeezing one online video platform into the business model of another, their proposal made startup founders across Europe worry about their future. Github, for example, is an open source code-sharing platform that helps developers to stay on top of trends. It too, would be within the scope of this law. Another example are crowdfunding platforms that, by design, host content uploaded by users. Is this where copyright infringement happens? Again, the scope of this law overreaches its aim and creates more problem than solutions.
Content industries may be yelling about a problem but it is lawmakers’ duty to find a proportionate solution for everyone in our society. In this case, technology can’t offer what politics wants. Several examples underline this: Shapeways, a 3D printing marketplace, hosts more than 300,000 pieces of copyrightable content per month but processed fewer than 1,000 copyright notices in 2016. Which were based on the description, not products themselves and in most cases unsubstantiated. Earlier examples of crowdfunding platforms, code-sharing platforms or e-commerce are no different. Because content recognition does not yet exist in an affordable and efficient way these startups can either break the law or break ties with Europe and move where common sense governs. Anyways, they are stuck between a rock and a hard place.
With audio files, a comparatively easily recognisable content, state of the art fingerprinting techniques resulted in error rates of 1–2%. Sounds acceptable? In comparison, spam filters for emails get dismissed as unsuitable with error rates of 0.1%. These cases illustrate: Filtering is ineffective with some types of content, and non-existent for others.
Even if filtering were to work properly across all formats, it would price many innovative ideas out of the European market. Studies have underlined this. Unlike an assessment by the European Commission suggested, filtering does not cost 900 Euro per month, but easily between 10.000 and 50.000. If the average initial funding of a startup was 150k, you can ask yourself whether you want to launch that company or just run your idea through a bad filter for three month.
Kickstarter, a crowdfunding platform, hosted 366.622 projects since its inception. In 2015 it received copyright infringement notices targeting a mere 215 projects, only one third of which were valid complaints. If article 13 of the proposed copyright directive became reality, the removal of 100 out of over 366.622 projects would easily cost 500.000€ annually. Why? Because some policy makers think a straightforward notice-and-takedown procedure is not enough. What was it again about proportionality in law?
When filtering is prescribed with complex and expensive rules, companies will be inclined to remove content rather than run the risk of getting sued. And who are platforms do decide if that video or drawing is a copyright infringement or parody, or maybe an entirely new work? While so far a well founded notice triggers removal, in future the benefit of the doubt will be with an armada of copyright trolls chasing anyone hosting content. The result will be less variety and content available online. This ranges from creative content of any kind to critical thought Here is a proposal that will lead to less investment, less startups and less free speech. Right before half of the world’s population will be able to benefit from a free internet, the most developed continent will go partly dark.
There are a myriad of startups like Kickstarter and Shapeways or Github. And even more young and talented Europeans are planning the next generation of content platforms today. The collateral damage of such out-of-touch legislation is not only a shot in the foot of Europe’s ambition to become a startup continent, but also a contribution to a generation of entrepreneurs seeking success elsewhere. Europe will be stuck with companies that are already big enough to comply or those who never want to be that big.
Startups are not one single industry but innovate across all sectors. They are the most mobile companies we’ve ever seen and are successful because they approach problems differently. Regulators are still catching up to this reality. While there is no simple answer to copyright, building walls will have unintended effects while missing the actual aim. Exempting startups, as suggested before will not crack the nut because startups aren’t SMEs. What then?
(This blog post first appeared in TheDigitalPost on 21 November 2017)