Intellectual property law to protect artificial intelligence systems against adversarial attacks in light of the purposes of intellectual property law
D. P. Bohatchuk · Juris Europensis Scientia · 2024
The paper analyzes whether the application of intellectual property law as a countermeasure against adversarial attacks on the artificial intelligence systems is consistent with the purposes of the intellectual property law.In view of rapid development of artificial intelligence and growing number of domains in which artificial intelligence systems are applied, the adversarial attacks on such systems can potentially cause very harmful effects.Thus, there is a need to identify and develop countermeasures against adversarial attacks, in particular with the help of law.The issue of the possible legal measures that can be applied to counter adversarial attacks has to be considered first of all in the light of the purposes of the law.This paper is dedicated to an overview of the purposes and theories of justification of intellectual property law, in particular, copyright and patent law, which are further considered in the context of artificial intelligence and adversarial attacks.Although the theoretical foundations underlying the intellectual property system is a subject matter of debates, the purpose of incentivizing creativity and innovation could be considered as the primary purpose of intellectual property law.The author concludes that the application of intellectual property law to protect AI systems against adversarial attacks and further integration of artificial intelligence into the intellectual property law system may be conducive to achieving the purposes of this system and implementing transparency.At the same time, modern intellectual property law should be adapted to the present-day realities, including the development of artificial intelligence.This paper can serve as a basis for further research into the possibilities and expediency of the application of intellectual property law to protect artificial intelligence systems and to counter adversarial attacks on artificial intelligence.