Why it is so important for companies to protect their algorithms

Today, algorithms are omnipresent in our everyday lives: they help us avoid traffic and automatically apply the brakes in an emergency, they recommend the right Christmas present for a loved one and make sure it is delivered in time, they check our spelling or help us find our soulmate. Dr. Katharina Scheja, Of-Counsel at Deloitte Legal and a lawyer specialized in complex IT projects and IP law, explains why – and more importantly how – we can protect these all-important algorithms.

This interview was conducted by Viola C. Didier, RES JURA editorial office.

Dr. Scheja, algorithms are – broadly speaking – just a set of precise steps. What makes them so valuable?

Dr. Scheja: “Yes, that’s right – they are essentially just a sequence of actions. Even something as simple as Cinderella’s rule: ‘The good ones go into the pot; the bad ones into your crop’ is an algorithm. We can also see from this famous example that algorithms are not necessarily tied to computers or big data, but rather a perfectly natural thing we all use in our daily lives. We know from playing chess, however, that it isn’t always so easy to define a specific sequence of actions, as there are literally millions of alternative options for every move. That’s what makes developing the algorithms used in chess software, for example, such a sophisticated undertaking. Solving complex tasks requires extremely complex algorithms and creating these algorithms not only takes an enormous amount of time, it also takes a creative, intelligent approach to succeed. So, it stands to reason that these complex algorithms have enormous value. Just consider the process known as Algorithmic Trading: the better an algorithm can calculate share performance, the more value it has – but obviously not if everyone else is using it.”

So, you are saying that algorithms are used everywhere and that they are essentially the lifeblood of the digital revolution. How can it be so problematic to protect them?

Dr. Scheja: “Well, it is pretty difficult to protect a simple sequence of actions. The results of an algorithm, say a certain software program, can be protected (by copyright). We have copyright to protect intellectual property and patents to protect technological inventions, but algorithms very often don’t fall into either of those categories.”

How, then, can enterprises effectively protect the algorithms they develop or those that are developed for them?

Dr. Scheja: “The only protection we have had to date is secrecy. As patent and copyright law were not available to us, it has always been – and still is – important to treat them as trade secrets and protect them through corresponding barriers to entry and exchange as well as through contractual agreements. That said, where confidentiality has been breached in the past, it has not always been easy to take legal action against an infringement and prevent the further use of the algorithm in question. EU legislators closed this loophole with the implementation of the EU directive on trade secrets and created a legal framework that applies across the entire territory of the European Union. That was a huge step forward.”

But don’t we have to protect our algorithms through internal measures as well – for example with regard to employees and service providers?

Dr. Scheja: “That’s correct – the lawmakers connected the two in the directive: it requires more effort in house, but also gives more rights to the owner of an algorithm. We have to implement and enforce confidentiality rules internally and adapt our contracts accordingly, whether they are with employees or customers and suppliers. The considerable effort involved in that really pays off, because we now have a legal framework to protect this valuable know-how – even on an international scale, for both the entire EU as well as for the US, where nearly identical laws were passed at the same time.”

Would you consider it negligence to not take advantage of the protection offered through the EU directive on trade secrets and the Trade Secrets Act which is soon to be passed in Germany? What is the worst case scenario?

Dr. Scheja: “To paraphrase the French diplomat and Napoleon advisor de Talleyrand-Périgord: ‘It would be worse than a crime, it would be a mistake,’ to shy away from the admittedly substantial administrative burden and relinquish your right to these new legal remedies. In the worst case, companies could lose access to their markets or even declare bankruptcy when key algorithms end up in the hands of competitors. We have seen scores of cases like these, with very few legal options available to plaintiffs. But now we have more options and companies must make use of them to protect their competitive advantage and improve or defend their market position!”

Your advice as an expert: what should companies do if they find their algorithms being used without authorization or copied?

Dr. Scheja: “First, don’t do anything rash, like perhaps give the alleged infringer advance warning! My advice: consult a lawyer specialized in this field and approach the case pragmatically and professionally.”

Many thanks for this interview!

Did you find this useful?