Algorithmic discrimination as a form of indirect discrimination is still unknown or poorly known to many in Serbia, but with the upcoming expansion of the use of artificial intelligence, there is no doubt that algorithmic discrimination will be on the rise and that from the very beginning one should bear in mind the necessity of strict regulation of it.
In order to be able to address the problem of algorithmic discrimination, that is, the negative consequences of the use of artificial intelligence, we must first know what artificial intelligence represents. Namely, artificial intelligence works by software based on a certain logarithm making decisions that most often relate to a large number of individuals. However, in certain situations, the decisions made by the aforementioned software, i.e. artificial intelligence, are not impartial and equal for everyone, which undoubtedly results in the existence of discrimination, which is precisely called algorithmic discrimination.
In March 2023, the Government of the Republic of Serbia issued a conclusion adopting the Ethical Guidelines for the Development, Application and Use of Reliable and Responsible Artificial Intelligence, which define different systems that will use artificial intelligence: systems related to biometric identification, systems that are intended for the management of road and transport traffic, systems that perform the selection of candidates in the education sector, etc. However, the implementation of these guidelines in the future may be problematic, precisely because the competent institutions do not have the capacity to act in the event of a violation of rights caused by the use of artificial intelligence.
So, for example, in March 2022, the state of California passed a special bill to combat algorithmic discrimination, which occurs as a result of the growing use of artificial intelligence. So, California has decided to tackle this problem head-on by passing a special law, which is likely to become a positive trend in other countries, both inside and outside the United States. This law regulates the situation by introducing certain legal obligations for deployers and developers, so that they are forced to pay more attention when creating a software for artificial intelligence, and the limits of their responsibility in the event of algorithmic discrimination are clearly set.
In our country, the situation is somewhat different, since the expansion of the use of artificial intelligence is still expected to a greater extent, which is why algorithmic discrimination in our country is regulated by existing regulations, primarily the Law on discrimination, the Law on consumer protection, the Law on the protection of personal data, etc.
Protection against algorithmic discrimination is also provided by international acts: the UN Universal Declaration of Human Rights, the European Convention on the Protection of Human Rights and Fundamental Freedoms, the EU Directive on the Implementation of the Principle of Equal Treatment, etc.
The increasing use of Chat GPT in recent months, even in our country, shows that the legal regulation of this area is extremely important, and that the application and effects of regulations adopted in this area abroad should be followed with great care. Just as California has taken measures to regulate the problem that arises as a result of the use of artificial intelligence, other countries are also expected to take a certain step in terms of more direct and precise regulation of algorithmic discrimination, which will most certainly represent one of the problems of the future for all. Although our country has not yet introduced any specific legislation regarding algorithmic discrimination, it can be said that it has taken certain steps in that direction by adopting guidelines, which is certainly a positive step.