The European Parliament’s rejection of the controversial Copyright Directive has highlighted the tensions in the evolving relationship between freedom and rights protection across the technology sector.
The ability of the law to effectively regulate developing technology has become a fiercely debated topic. The proposed directive is the latest indication that legislators are falling behind in ensuring adequate protection for intellectual property (IP) rights holders, particularly given the increasingly broad applications and uses of artificial intelligence (AI).
Given the obvious benefits of IP protection and the fact that legislation needs to catch up, why is the Copyright Directive so controversial? The primary concerns voiced by many, including a number of internet service providers, were with two particular aspects of the directive – articles 11 and 13.
Article 11 proposed the introduction of something now commonly referred to as the “link-tax” which, if implemented, would have required online platforms hosting links to news sites to pay a licence fee. The aim was to try to ensure that news sites benefited from third parties linking to them, rather than the other way around.
In a similar vein, Article 13 would have required websites and online platforms to take measures to ensure they obtained appropriate licences from IP rights holders to host or show any content uploaded to their sites, whether by the creator directly or by third parties.
IP for the internet age
The theory behind the directive in general, and these articles in particular, is to update IP protection for the internet age by protecting rights owners, and supporting artists and public information organisations from copyright infringement.
Many celebrity artists, including Paul McCartney and Annie Lennox, have been vocal in their support for the bill, principally on the basis that it would ensure creators are compensated, or better compensated, for any use of their work. Many artists argue this frequently does not happen under the current legislation.
While the motivation for the directive seems sound, with creators being provided with greater scope to protect their rights, the objections largely stem from the predicted cost and massive undertaking that enforcing the legislation would involve, when many are convinced that the practicalities of enforcing it are unworkable.
An obvious example of an entity which would be affected by the directive, if it were implemented in its current form, is Wikipedia. Italian Wikipedia felt so strongly about the potential effects of the legislation that it shut down for a day in protest, citing the risk of closing if the legislation were to be passed.
The general consensus is that smaller online platforms would be at particular risk of being forced out of the market following implementation of the directive, if it went ahead. Big players such as Google and Facebook are likely to be in the minority of online platforms that could weather the potentially significant licensing costs.
Means of enforcement
The other major challenge foreseen with the proposed changes in the directive was how they could be enforced. Reportedly, there are 400 hours of video content uploaded to YouTube every minute. With internet content being produced at this rate, the sheer amount of surveillance required to continually monitor for copyright and/or other infringements of intellectual property would be immense.
There is no question of humans having the capacity to monitor new and existing data on such a scale. For internet-wide protection of IP rights to be a possibility, whether under a revised draft of the directive or otherwise in the future, an alternative means of monitoring for infringements would need to be found. Not least because protective rights are of very limited use if there is no practical means of enforcing them.
Many argue that automated content monitoring systems impact negatively on creativity and development. AI has developed rapidly in recent years, but still lacks the sophistication of the human reasoning process. Existing copyright monitoring systems used by sites such as YouTube are well known for errors in banning acceptable content due to algorithmic inabilities to differentiate from infringing content.
Internet leaders, including World Wide Web inventor Tim Berners-Lee and Wikipedia founder Jimmy Wales, have expressed significant concerns regarding the capability of AI to be used for IP protection. In particular, they highlighted concerns in relation to the status of content recognition software. As with all forms of AI, there are challenges as to who has liability for any losses incurred when AI gets it wrong – which is unavoidable.
Looking wider than just copyright infringement, Jim Killock, executive director of the Open Rights Group, raised the point that intellectual property courts spend hours deliberating whether or not cases amount to infringement of rights. It seems impossible for an automated system to be nuanced enough to do the same thing effectively.
Although the directive has been rejected for now, there is no doubt that the current legislation needs updating to better protect rights online. For as long as technology and AI continue to grow, the law will have to grow with them to ensure that existing rights remain protected.
AI is becoming increasingly integrated into a number of sectors, with improvements in interfacing allowing for better human/machine integration and, accordingly, increased productivity. Advances in AI will change the modern workplace over the next few years and it is clear that the pace of legislative changes and the means of enforcing these will need to dramatically increase to keep up with this.
It seems likely that steps will also need to be taken by those creating and amending the existing law to incentivise the creation and innovation of appropriate systems, AI or otherwise, to allow for the protection of IP so that it continues to flourish.
Read more about digital copyright