Artificial intelligence, or ‘AI’, is undoubtedly a hot topic in any industry at the present moment, with corporate spending on AI estimated to have been around $50 billion in 2020, despite most company managers not reaping many benefits from their investments. Therefore, there are two significant things which need to be considered with these developments, the first being how AI can be incorporated into the legal field, and secondly how this can be governed, controlled and understood, particularly where AI creates things capable of being protected by intellectual property rights.
How is AI used in the legal sector?
One of the most tedious activities for lawyers is undoubtedly editing, amending and checking contracts, which can be hundreds or thousands of pages long. In addition to this, this task could also be clouded by human error, which is why there is a market for a development of technology which can make this process easier. Clearlaw is one of the artificial intelligence systems invented for this purpose.
AI in terms of legal research is probably the most popular form, and most well-known and widely used. Databases such as Westlaw, LexisLibrary and Practical Law are used in most if not all law firms and universities to enable research to be conducted efficiently. Using these platforms, one can easily search for key words, key cases or phrases from legislation. This gives the user a lot of spare time, which would be spent scanning through books to search for one relevant part of information. However, these systems mentioned are not very advanced in the intelligence aspect, as they can only search for words and phrases as you put them in. This is where new, more advanced, systems such as CARA AI from Casetext have come into the playing field, having developed more advanced software to find cases specific to the points of law you need.
AI has also been used by the London Metropolitan Police in cameras which use facial recognition and scan ‘watchlists’ as a mechanism to help prevent and fight violent crime. It works so as to recognise faces of known serious offenders, which then allows police officers to approach and arrest them if they are confirmed to be a known suspect or person of interest. The first arrest using this system was successfully made in February 2020, however, there are clearly privacy concerns and also concerns with regard to unfair data profiling of these individuals. The reconciliation of personal privacy including data protection and national security is a big task in this respect.
Regulation of AI
A big question with regards to AI is whether it has its own separate legal personality, like that of a company. There is a problem with this idea however where if AI did something illegal or infringing, the AI itself would be liable, and of course, a machine cannot physically be punished or put in custody. In addition, if an autonomous vehicle is involved in an accident, there becomes an issue as to who would be liable for it. Would it be the inventor of the software? Alternatively, the manufacturer of the car?
This is a particular issue where intellectual property rights are considered, purely because if a human inventor makes AI software which is then capable of creating its own invention, there is an issue as to who can be deemed to be the ‘author’ of such works. This is an important question because the author of any work will be the one with whom the benefit of an intellectual property right rests. The Copyright, Designs and Patents Act 1988 sheds some light on this in s.9(3) which states:
However, this only seems to clarify the position on authorship in works capable of being copyrighted, and not those subject to other rights.
This, with copyright specifically, shows how difficult it will be to actually fully regulate AI. The scope of the necessary regulations is so wide that it cannot simply be covered by a wide statute stating everything AI can do, could be liable for or the control of it. As there are so many applications of the technology, such as the ones in the legal sector named above, there can be no comprehensive law to cover it all – especially if the laws that regulate autonomous vehicles are the same as those regulating a computer programme. To attempt to do this would be trying to control something uncontrollable.
Conclusion
AI is an advanced software that can be used for many different inventions, both in the legal sector and in others. What is so interesting about its use in the legal sector, is that it will not only save time for lawyers who decide to use it, as in the ways mentioned above, but as a result will reduce costs for clients due to billing hours being lessened.
In addition, the legal sector has always been a sector which has been labelled ‘traditional’ and therefore there is the assumption that lawyers are reluctant to use legal tech in the profession. With the growing popularity of AI in all aspects of business, lawyers also need to be aware of the regulation of it because it would be incredibly difficult to advise on an AI created design or product without having a proper understanding of the authors and consequent rights that are afforded to it.
As with all new inventions, products and information, it will take time to understand a clear picture as to where AI will fit into other aspects of law, especially with regards to liability of the machines. As statutes, cases and reports begin to cover these new territories, we will all have a better understanding of how AI can be regulated effectively and used by all within the legal profession.
Comments