This article discusses further an aspect of Agnieszka Kraińska’s article on legal personality and artificial intelligence. I take a slightly perspective of the subject. Even today new categories of legal entity are needed, unrelated to AI systems.
Even at present there are technologies for which a new type of legal entity is needed. This is blockchain technology, in particular smart contracts. For the sake of this discussion, a few simple examples of smart contracts are usually given, for example a smart contract that creates a gambling code. This is a relatively simple code, which even today can successfully be implemented on a blockchain. Players make a payment of the specified amount of the cryptocurrency to the smart contract address, and the smart contract performs a draw of the winning address at regular intervals (this is one of the addresses from which payments were made) and transfers the prize in full to that address.
This simple smart contract presents a major challenge to the current legal system. At this point I am not concerned whether or not this is legal in a particular legal system. If we assume for a moment that it creates some kind of legal relationship, it needs to be considered which entities are party to the relationship, and which entity the winner of the lottery should specify as the source of revenue. As the legal system does not at present recognise legal personality of a smart contract, it is transparent from the point of view of the law. Therefore, formally, if some kind of legal relationship is formed in the case of the smart contract in question, then it is between the lottery participants. The legal relationship is therefore multilateral and the parties are anonymous entities scattered around the world.
Intuition tells us that this approach is not entirely in line with the way in which the smart contract participants in fact see this legal occurrence. They have the impression above all that they enter into a relationship with the smart contract. Even if this is not the case and in fact they feel that they are party to a multilateral legal relationship, this is an impractical approach in every respect. Even if it is assumed for administrative-legal purposes (for example for the purpose of tax law) that the winner’s revenue was generated under a multilateral legal relationship, this may give rise to a number of major challenges (for example identification of the parties to the legal relationship).
Due to similar considerations, decisions were made in the past to introduce legal personality for human associations and estates, creating the concept of legal persons. These are abstract entities (like smart contracts) in which a range of practical problems related to multilateral social and economic relationships can be eliminated. Introduction of a legally relevant entity between the end participants of multilateral relationships simplified trade in a very effective manner and allowed entirely new economic models to develop.
The same applies to smart contracts. Like blockchain, they form the potential framework for an entirely new dimension of the economy. Without recognising the legal personality of smart contracts, the potential that this technology creates may remain untapped. For the purpose of this article I limited this analysis of smart contracts to a simple gambling contract. We need to be aware however that smart contracts could be used to create significantly more advanced economic structures (insurance contracts, decentralised organisations, etc.).
Someone may wonder what smart contracts have to do with AI and why smart contracts are being discussed in the context of the proposed legal personality for artificial intelligence. One of the most fascinating technological trends in the next few years will be attempts to integrate AI technology and smart contracts.
Once these two types of technology are combined, there will be not only a smart contract that is distinctive (even today there are many cases of this). The decision-making processes in it will also largely be autonomous. Before long, smart contracts might be integrated with autonomous algorithms (black boxes). In such a case, transactions initiated by a smart contract will be a result not only of a predefined code, but also to some extent of autonomous activities of an algorithm. This is an area in which ideas are being developed for example to create investment funds on a blockchain. They will take the form of a smart contract that is highly autonomous in its decision-making.
The Polish legal system is not ready for solutions of this kind. When attempting to adapt the known legal structures to them, we encounter many difficult problems or create artificial and impractical solutions. Only recognising legal personality of these structures can open new possibilities. It is time to prepare for this.
Krzysztof Wojdyło