| With the development of artificial intelligence technology,artificial intelligence technology is subtly changing the way of daily life of the public,and the artificial intelligence algorithm technology hidden behind it has also received more and more attention.The number of patent applications is increasing year by year,and more and more artificial intelligence algorithm technologies are eager to be protected by the patent system.However,for the intellectual property protection of artificial intelligence algorithms,China’s Patent Law does not give clear provisions.Only in the Patent Examination Guidelines are sporadic provisions.However,the nature of the Patent Application Guide is not law,but only a guide.Artificial intelligence algorithm artificial intelligence algorithm is the soul of artificial intelligence technology,at the technical level,artificial intelligence algorithm is at the bottom,so it is not easy to detect.But this does not mean that it has no protection value,on the contrary,it is precisely because it is at the bottom of the technology,it plays a vital role in the development of any artificial intelligence technology.This paper combs through the relevant literature in recent years and finds that although at the theoretical level,there has been a call for patent law protection for artificial intelligence algorithms,but the patent law has not been clearly stipulated,so there has been controversy.Therefore,the innovation of this paper lies in combining theory and practice,analyzing the theoretical basis of the patentability of artificial intelligence algorithms on the basis of previous research,and at the same time summarizing the conclusion that artificial intelligence algorithms must be combined with specific technologies if they want to obtain patents through practical cases,and put forward targeted and effective suggestions under the condition that they do not violate the current patent law.The disadvantage is that the theoretical argument process for patentability is not very perfect,slightly brief and inaccurate,some of the arguments and arguments are not convincing,and the language is not refined enough.The distinction between relevant subject matter is still slightly vague and lacks practical experience,and the relevant suggestions put forward may not be feasible and practical.This article is mainly composed of the following parts.The first part starts from the basic concept of artificial intelligence algorithms,gives a clear definition of artificial intelligence algorithms,and classifies artificial intelligence algorithms from both academic and technical perspectives.The technical characteristics of artificial intelligence algorithms are summarized.According to these technical characteristics,the AI algorithm is clearly distinguished from other similar objects.The second part focuses on analyzing the patentability of AI algorithms,analyzing the theoretical basis of patentability of AI algorithms from the external form and substance.This paper argues that AI algorithms are not only basically the same as traditional objects in external form,but also meet the patent eligibility criteria in substance,so there is a theoretical basis for granting patent protection to AI algorithms.At the same time,the situation of artificial intelligence algorithms in the United States and Europe was sorted out,and it was concluded that both sides believed that artificial intelligence algorithms were worth protecting,but the focus of the review was different,the United States focused on the technical attributes of algorithms,and Europe focused on thinking attributes.This is of great significance for China when reviewing the patentability of artificial intelligence algorithms.This paper uses the Amap mobile phone software to predict the time of traffic lights through the algorithm,pointing out that as long as the artificial intelligence algorithm is not a simple expression of mathematical ideas,and has technical properties,it can be patented.The third part focuses on the challenges faced by AI algorithms,mainly including legal barriers to the patentability of AI algorithms and some other issues.Among them,the legal obstacles are mainly that artificial intelligence algorithms need to face the examination of "patent three natures".This paper argues that the practicality,novelty and challenges of AI algorithms in inventive step examination are closely related to their special technical attributes,such as the close integration of AI algorithms and big data,which makes it difficult to define the specific scope of the term "art" in inventive step examination.But just because AI algorithms pose challenges doesn’t negate their patentability.The fourth part puts forward suggestions for improvement,and proposes corresponding and targeted measures for different challenges,including the refinement of the patent subject matter exception rules,and the full protection of artificial intelligence algorithms under the framework of the current patent law. |
