Artificial Intelligence refers to the ability of machines to exhibit intelligence, performing cognitive human functions, like learning, problem solving and enhancing their abilities without any sort of human intervention. Virtual personal assistants like Siri and Google Now, self driving cars, search engines, image recognition, etc., are leading examples.
AI consists of a built in algorithm that allows it to learn from the data inputted and to make decisions. When applied to art, literary and musical works, new and independent units of work are created without or very minimal human intervention. While the programmers have the power to set parameters, the work or final output is generated by the AI itself.
A recent example of such programs which are completely independent is the painting “The Next Rembrandt”, produced in the Netherlands in 2016. Other AI projects include a 2016 Japanese program which produced a short novel and Google’s “Deep Mind” program.
So, the pricking legal problem which the government has to address is that when an artificial intelligence creates or produces any original work on its own, is it entitled to a copyright over its creation?
One possibility is to let such pieces of work be free from any copyright because they are not created by a human author, being freely used and reused by anyone. But, it would result in a big disadvantage to the companies investing and selling such creations as they would be used freely without payment as they are not protected by law, leading to huge losses for companies and an unwilling to invest in such projects anymore.
The Indian legal system does not have any direct provisions to address the problem of copyright by AI generated works. Generally, the inventors or creators are granted patents and copyrights on the basis of the fulfillment of certain criteria like creativity, inventive steps, novelty, etc. But, nothing is mentioned about innovation by non- human entities, leaving the question of ownership of creations by AI still unanswered.
Presently, AI programs are not considered as a person and so, there exists the problem of providing copyrights and patents on the basis of ownership of the content. Under Section 2(d) of the Copyright Act, 1957, ‘author’ is defined which has given rise to certain difficulties.
It is essential to reflect on the overall social benefit achieved by providing copyrights to such works, whether the society would benefit most if the copyright is assigned to the AI programmer, the organization funding the development of the AI, or the potential end users of AI programs. If an author cannot be determined clearly, one of the available options is to allocate the authorship of the work to the human or legal person nearest to the creation of the produced work, even if that person does not ‘cause’ the work to be created. Another option is to declare the work as having no author or owner at all.
The option of allocating authorship to a human or legal entity is the most popular and practiced currently even though it ignores the contribution of the AI completely, considering it only as a tool. Contrary to it, if the AI lacks any legal personality, the option of not awarding ownership at all appears to be more appealing. Though, the situation changes considerably if the AI is recognized to possess a separate legal personality or holds the status of ‘electronic persons’.
This would help to add more flexibility and freedom in the public domain. But, a major issue is that AI programs are not so independent and do rely on human input to a certain extent. Thus, hampering incentives for companies and investors to produce such programs in future, thereby, reducing the commercial viability of created work.
Independent programmers may be allowed to retain copyright for the work produced by their AI programs. But, for large companies, the authorship may be settled through employment contracts and credited to the programmers or the companies for which they work on. In the long run, using licensing may prove to be more financially viable for companies as well as aid in the commercialization of AI generated works for the benefit of the society as a whole.
There are active indications that the laws of a majority of countries do not and will not easily recognize or accept non- human copyright like in USA and in Europe’s Court of Justice of the European Union (CJEU), where it has been declared that copyright only applies to works that are original and that reflect the author’s own ‘intellectual creation.’
In India, the option of giving authorship to the programmer is present which is similar other countries like Hong Kong, Ireland, New Zealand and the UK.
Thus, there is an essential need to provide the proper financial incentives to encourage the growth and development of the AI industry and to ensure the distribution of works generated by AI as currently it is humans who hold the copyrights for such works. AI programs and machines do not require financial incentives. Their performance is dependent on the investment of time, money and skills by programmers, investors and companies. Their contributions are essential, as without their research and development, AI programs would not be available for public use like it is today.
An effective solution is yet required to address these differences between the technological and human aspects of the legal status to provide copyrights which should be both motivational to developers of AI programs and should not disrupt the current legal system. Therefore, the question of the law pertaining to the ownership of work between humans and AI is bound to get more and more confusing in the years to come as the differences between human acts and AI would always remain. Thus, there yet remains a need to understand, create and develop laws accordingly to undertake the new and upcoming role of AI in the day to day life.