The year 2022 marked a fresh and innovative beginning for the field of artificial intelligence (AI), wherein the technology was re-introduced in a more accessible and usable form to the public. While the roots of AI date way back to the era when Alan Turing's machine decoded the German code ‘enigma', it is needful to mention that the developments in the field since that war have been manifold. The most recent development was the introduction of a downloadable web application which acted as a chat-bot for people who were either curious about or in need of information.
On November 30, 2022, a company called Open AI announced the launch of this unique search engine known as ‘Chat GPT' or the ‘Chat Generative Pre-trained Transformer'. This artificially intelligent chat-bot was designed to provide outputs in a human-like manner.
While the application followed the basic concept of a generic search engine like that of Google or Bing, it, however, promised a unique interface where instead of the user-facing an endless list of possible solutions to their queries, the algorithmic search engine would provide them with a simple answer or solution, thus portraying a somewhat simplified user interface. Moreover, the accessibility of ChatGPT was accompanied by a vast database of knowledge, through which the application was able to solve some of the most complicated equations, which even led to its successful passing of final examinations for the MBA course posed to it by a professor from the Wharton School of Business, thus implying that the application could have attained the degree without any hassle. Not only this, but the technology was also capable of providing users with well-developed professional emails, essays and other literature which would otherwise require an application of mind.
Owing to such a proven genius of the App, certain users sought to reignite the concerns pertaining to the ways in which ChatGPT was accessing and producing this information alongwith the amount of personal information that the bot was accessing to bring the ‘best possible' outputs for a particular user.
Needless to mention that these concerns, being relatively lesser as compared to the popularity of the application, many individuals were easily volunteering information to ChatGPT for various purposes ranging from the need to attain answers to the testing of the potential of the application. Recent usage of the application which was widely published on different social media applications revealed that individuals added information relating to their employers, college professors, assignments and even clients, with the objective of having the application draft emails, research papers, examination answers and various other documents which may otherwise be designated as confidential and would have raised issues pertaining to a breach of non-disclosure agreements in certain cases.
Although it is not presumed that such practice will continue in the future, however, it may be appropriate to assume that volunteering of certain classified data to the application for appropriate answers will always be a possibility. Furthermore, it is also needful to note that one of the primary features of the application is that it studies the behavior of the user to provide them with the most appropriate output, which further raises question pertaining to the extent to which the application can access personal information of the individuals to learn their behavior or pattern.
Right to Privacy and Social Media
Social media usage has always been plagued by gross misuse of users' personal information by leading companies. Conglomerates like Meta Inc, Twitter Inc and even ABC Inc have faced international and national scrutiny on their illicit use of such information for monetary and popularity gains. This abuse has increased to such an extent that these companies have even been periodically summoned by the United States Senate to face congressional hearings.
Even the Competition Commission of India (CCI) initiated suo moto proceedings against Meta Inc for abusing its dominance over the market by updating the terms of use of its messaging application ‘Whatsapp' and allowing the sharing of user data across Facebook Inc. While the CCI in this case titled ‘In Re: Updated Terms of Service and Privacy Policy for Whatsapp Users1', held that WhatsApp was in contravention of the law, the latter company challenged the order before the Hon'ble High Court of Delhi, and subsequently before the Hon'ble Supreme Court of India2, wherein, both adjudicatory bodies upheld the decision of the CCI. Recently, in yet another attempt to fight for the legal validity of its privacy policy, Meta Inc. in the case of Karmanya Singh Sareen & Anr. vs. Union of India & Ors3. claimed that owing to the pendency of the Draft Data Protection Bill, the judgment of the Hon'ble Supreme Court in the 2016 case must be put on hold. The matter, being sub judice, is yet to determine the fate of the data possessed by the people of India.
Nevertheless, it is manifest to note that the debate pertaining to privacy is ever-present in the social media hemisphere and is deemed to be ongoing due to the increasing overreach by the conglomerates, accompanied by the absence of a clear definition of the manner and the lengths to which the information volunteered or acquired from the user is utilized.
To add to the icing, ChatGPT, being an artificially intelligent application capable of optimally using personal information for the benefit of the application and detriment to the user, and where the privacy policy calls for ‘limitless' usage of user information, makes it evident that the threat of privacy infringement may increase to a substantial degree.
Reading of Privacy Policy
Within the privacy policy of Chat GPT, which was last updated on September 19, 2022, it is specified that Open AI may share the personal information of the user with third parties without prior notice, unless such notice is required under law. It is further provided that the sharing of information may also be without limitation in cases where the same is being shared with Vendors and Service providers to assist Open AI in meeting ‘business operation needs' and to perform ‘certain services and functions'. However provided by the fact that the company does not define the terms ‘business operation needs or ‘certain services and functions', the extent and reasons of the data shared becomes ambiguous. Moreover, provided by the prima facie fact of Chat GPT being an AI, it is probable that such usage of personal data is optimized to the best benefit of the company and may be more detrimental to the user than when the same is being shared by another non-AI company like Meta Inc.
It is, however needful to note that this potential does not preclude the latter companies from being able to limitlessly and optimally use data as even Meta and ABC Inc. have introduced their artificially intelligent applications, namely the ‘Meta AI' and the ‘Bard', which may have similar privacy policies.
The Chat GPT privacy policy further specifies that personal information of the user may also be used for ‘conducting research' which may either be internal or shared by third parties or even published or made generally available. A mere perusal of this clause provides a clear inference of a considerable increase in the ambit of public sharing of information. By including the term of publishing or making any information generally available, coupled with the fact that the company has called for a limitless approach to information in undefined circumstances, users may have to exercise additional caution before using the application.
The need of such caution is furthered by the fact of the policy calling for a collection of data concerning the use of services, type of content viewed, device information and other relevant information of the user.
Statutory Infringement and Usage Policy
The question pertaining to the infringement of copyrights or other intellectual property in the case of the use of Chat GPT is two faceted. It is undeniable that the artificially intelligent application has to attain the needed information from some source, thus any usage of such gathered information may, in turn, lead to the user being prone to an infringement suit.
This fact is further proven by clause 3 of the Terms of Use published by Open AI, which specifies that owing to the nature of machine learning, the output may not be ‘unique' across users. This characteristic is essential for a copyrightable work. It is also essential to note that the company is well aware of the possibility of such infringement and has provided a clause stating that upon being notified about any purported infringement, it may delete or disable the content in question.
Another aspect of intellectual property issues is the right of Chat GPT. The information originally sourced by the chat-bot has been specifically mentioned to be a property of Open – AI, the illicit usage of which may again open the user to infringement actions by the company. While the policy does not specifically provide for clauses pertaining to copyright infringement of Chat GPT, it restricts the user from using the services in a way that may infringe, misappropriate or violate the rights of any person. Another peculiar term of usage seeks to restrict the user from representing the output, which in this case would be an answer, was human-generated when it is not.
These requirements give a reasonable inference to the fact that, wherein the users have been utilizing the outputs for writing their essays or even work emails, they may be prone to legal action by the company, in line with the policy.
Contrary to the policy, Indian and majority of other foreign legislations are yet to provide protection to artificial intelligence or other self-generated computer programs under the law. In India, while the Copyrights Act, 1957 provides for specific provisions pertaining to the protection of computer program separate from the ones for the protection of original literary, artistic, musical, or dramatic works, however, a provision pertaining to the protection of the literary, artistic, musical, or dramatic works, created by a computer program has yet to be included. In fact, under certain jurisdictions, including that of the United States of America, a copyright can only be registered by a human being.
Even otherwise, the universally accepted works eligible for ‘copyright protection' in line with the definition given by the World Intellectual Property Organization include literary and artistic creations, such as books, music, paintings, sculptures, films and technology-based works (such as computer programs and electronic databases), wherein such technology-based works are ones which may be created by human beings.
Conclusion
As previously mentioned, the advent of artificial intelligence is not new to the world, wherein different companies or conglomerates continue to work towards making the technology more economically feasible and user accessible. With the launch of Chat – GPT, this fact of development and the growing popularity of AI in technological advancements has been further proven. In fact, even various law firms have started initiating the process of introducing artificial intelligence systems within their functioning, an example of which is ‘Harvey' a generative AI developed specifically for complicated legal issues faced by law firms in their cases.
Even during the Union Budget 2023, the Government of India resolved to establish three departments in leading educational institutions for conducting research and development-related work in the field of Artificial Intelligence. Thus, it is apparent that the future will only witness more ways in which technology is brought into the world.
Chat – GPT has proved to be a revolutionary application by bringing the features of a search engine in a chat bot. Needless to say, the functions provided and promised by it are far greater than any search engine developed by a company. While the application has had various predecessors developed by other companies and given different human like names like Eliza, Alice, Jabberwacky and Mitsuku, neither of these applications have gained the kind of approval as that attained by Chat – GPT.
It is only through this popularity that the debate relating to the use and misuse of data by different web based applications has regained momentum, which has further provided for a dire need of the development as well as an amendment to laws so that any legal infringement by these applications can be aptly dealt with. The extent to which the Draft Data Protection Bill, 2022, will be able to control the use of the personal information of an individual, is difficult to determine, until its complete implementation in India is observed, however, slight amendments to the Intellectual Property Laws, may lead to the protection of copyrights and trademarks for whichever side, as may be applicable.
Even though the statement of Turing specifying how machines tend to take him by surprise with great frequency may have been given in a different context altogether, it is essential to note that users may need to be sufficiently satisfied and well aware of all policies of these applications before using them in order for this surprise to not turn into devastation.
Footnotes
1. In Re: Updated Terms of Service and Privacy Policy for Whatsapp Users, Suo Moto Case No. 1 of 2021.
2. WhatsApp LLC & Anr. v Competition Commission of India, LPA 163/2021 & CM APPLs. 15908/2021, 16893/2021, 18800/2021, 18910/2021, 46058/2021, 46059/2021, 46655/2021.
3. Karmanya Singh Sareen & Anr. vs. Union of India & Ors, SLP (C) 804/2017.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.