Big data, big problem?

Speakers at the recent joint UK IPO and WIPO Conference on artificial intelligence and intellectual property (AI: decoding IP) explained that having the best AI algorithm may not necessarily be a good thing for the company who owns the technology. While having masses of data to train AI and improve the output of the algorithm is good from a technology standpoint, it may cause companies a massive headache from a competition law perspective. In this article, we briefly look at the issues that in-house attorneys may need to consider when dealing with AI.

Competition law in the UK and EU (and anti-trust law in the US) exists to ensure businesses are competing on a level playing field and are protected from others acting unfairly. However, technologies that use vast quantities of data, such as artificial intelligence and big data processing, are tipping the playing field. 

Abuse of Dominance

One issue is the existence of key gate keepers. Typically, larger companies with technology platforms that everyone needs or uses also hold large volumes of data. The first company to operate in the space may have time to acquire the most data, and potentially gain a dominant position in the market. However, this may inadvertently make it harder for new businesses to enter the market despite the cost of starting a software business being relatively low (compared to, for example, biotech).

Owning, and not sharing, a large amount of data may exclude others from developing a competing technology. Some companies with large amounts of data may decide to share or pool their data with selected partners, since more data generally means better AI results. However, the arrangement may prevent others from accessing the data pool and thereby entering or retaining their position in the market. Exclusive licenses to technologies or exclusive access to data sets or a data pool may be considered an abuse of a dominant position.

However, in some cases, users of the AI technology may prefer all their data to be in the hands of one company. For example, where the AI is being used to provide a healthcare app (e.g. an AI-based doctor app for assessing symptoms), users may prefer that their personal medical data is held with a single company rather than shared with many. Therefore, the type of data, privacy and data protection may be relevant when considering whether the size of a data set owned by a single company is anti-competitive.

AI-Based Pricing Strategies

Dynamic pricing strategies that use AI could analyse large amounts of data, identify local or global buying trends and determine more competitive prices for products, which benefit us as consumers. AI algorithms could also provide individual pricing strategies for consumers, based on what individuals spend money on, how responsive they are to offers and limited-time deals, whether they have loyalty cards, and so on. However, problems may arise when a company outsources their pricing problem to an AI that outputs dynamic pricing strategies. What if the AI determines that the best way for the company to maximise profit is to not engage in price wars and to instead fix a price? What happens if multiple companies outsource their pricing problems to the same AI supplier? 

Merger Control in Digital Markets

Another issue that may arise is when company A acquires company B (and its data) and, as a result, company A now has a very large data set. The acquisition of company B’s data enables company A’s AI to be improved, but that data is now not available to anyone else. Mergers are common in the software space, particularly since companies tend to specialise in either developing the algorithm or acquiring the data for the algorithm, or because one company has the innovation and another the resources to commercialise the innovation. Therefore, there is a need to tread carefully and consider whether the data acquired as a result of a possible merger may negatively impact other players in or new entrants to the market.  

There was some discussion at the conference on whether competition law needs to be updated to reflect the new digital economy and associated and new technologies, or if a different regulatory regime may be required entirely to regulate the gate keepers, or if anti-competitive behaviour in this arena should be assessed in view of other issues, such as privacy and data protection.

To find out more about the European Commission’s thoughts on competition law in the digital age, please see this recent report

At Appleyard Lees we have a large and experienced team of attorneys and solicitors who combine a broad range of legal and technological expertise with real world commercial insight. If you need help navigating the complex world of intellectual property for AI innovation, or want to know how any of the above issues may impact your commercial strategy, please contact us.

Author

Related posts