ARTIFICIAL INTELLIGENCE: COPYRIGHT CONCERNS V. TECHNOLOGY DEVELOPMENT BY - MR. HARSHA C & MS. AKHEKIVIMI K CHISHI
ARTIFICIAL INTELLIGENCE: COPYRIGHT CONCERNS V. TECHNOLOGY
DEVELOPMENT
AUTHORED
BY - MR. HARSHA C
2nd
Semester, LLM in Intellectual Property and Trade Law at
Christ
University, Bangalore
CO
– AUTHOR - MS. AKHEKIVIMI K CHISHI
2nd
Semester, LLM in Constitutional and Administrative Law at
Christ
University, Bangalore
ABSTRACT
Artificial
intelligence (AI) creates content in the form of literature, art and music
which is protected under copyright. The TRIPS agreement not specifying about AI
generated works and there are different approaches in domestic legislation,
requiring human creativity that effectively leaves such works unprotected. AI
regulation is linked to questions of morality, and intellectual property. The
AI content has an issue of authorship and ownership issues and there are
challenges in the development of AI that international trade rules could
address, such as improving global access to data to train AI systems which is
copyright infringement and there is lacking of clarity as to the liability of
such content creation and misinformation and defamatory information. With this
lack of clarity about authorship and liability, there is no scope for the
development of Artificial intelligence-oriented services and technology as
there is no access to content and scope for infringement. This article tries to
bring clarity about who should own the authorship and liability and how can the
copyrighted content be used to improve artificial intelligence-oriented
technology and services in an international platform without any infringement
claims.
Keywords:
Artificial Intelligence, Copyright infringement, authorship, liability and fair
use.
1.
INTRODUCTION
Artificial
Intelligence, everyone talks about it and uses it for daily needs to ease their
work such as generating information on any topic, draft documents, mails, etc
and there’s likely no problem that anyone has faced when using artificial
intelligence in the rough sense that the information is used only as a draft
and to understand any topic or use it for non-commercial and personal use.
Artificial
intelligence? How intelligent is it, what does it do and how is it different
from other information available on any search engines like Google, Bing, etc.
Before anyone can think of the legitimacy of the content generated by artificial
intelligence or artificial intelligence in total, it becomes important to
understand the use and the need for artificial in a general sense in the
perspective of the common man.
Before
artificial intelligence, search engines like Google, etc have taken people's
attention to get information and many other things such as acting as a bridge
between other sites and companies and the people. Any user who is subjected to
the internet and browsing before artificial intelligence would first look for a
search engine like Google and then type their query in the search box and then
redirect themselves to the relevant sites.
What
is the status and need of such bridging sites like Google to connect people to
other sites when the sites already have their website links and can access them
directly? The search engine has got people's trust and they believe whatever
their query is, can be asked in the search box of such search engines and they
get relevant information in no time either to be redirected to other sites or
get content. There was no problem regarding the status of such search engines
as they only act as a platform to provide information available on other sites.
What
is the problem with artificial intelligence in generating information when
there is no problem with search engines providing content? Is it because it
doesn’t provide a direct reference to the content generated or is it the
validity of the information generated?
From
a common man's perspective or need, what difference does it make? People used
to use search engines to get information and now use artificial intelligence to
get information, draft documents, emails, etc. With the evolution of
technology, the ease of work too has improved either with the evolution of
search engines or artificial intelligence. So, what’s the problem with
artificial intelligence generating information?
To
understand this, we need to understand the definition of artificial
intelligence, how it generates information and how different is the information
generated by search engines or other platforms to that of artificial
intelligence and how is it related to intellectual property specifically to
copyright.
2.
ARTIFICIAL INTELLIGENCE – A RESEARCH APPROACH
Artificial
intelligence as simple as it makes people’s lives is the complexity of its
nature and definitions. Artificial intelligence is defined by many researchers
and some of them are as below for the understanding of the meaning and concept
of artificial intelligence.
Samoili
S and López Cobo, et al.[1] defined artificial
intelligence[2] in their
article “AI Watch. Defining Artificial
Intelligence. Towards an operational definition and taxonomy of artificial
intelligence” (2020). WIPO[3] has
defined artificial intelligence in the “Revised
Issues Paper on Intellectual Property Policy and Artificial Intelligence”
(2020)[4].
There is no universally accepted definition of artificial
intelligence. Every country or
organization has its understanding and stance on artificial intelligence making
it difficult to interpret what is artificial intelligence.
As we have seen, from a common man's perspective and the rough use
of artificial intelligence has no exact problem as such but is it really having
no problem in total? This article tries to identify the complexity and
challenges for the authorship and ownership of the content and artificial
intelligence as a separate entity in the realm of copyright and understanding
the liability in generating content and other issues related to the artificial
intelligence.
As we have tried to compare artificial intelligence to search
engines in the introduction of this article, we can compare both from the
researcher’s perspective to understand the authorship and liability issues of
artificial intelligence generating content. What is different between
generating the content and users using any such content?
Let us illustrate[5] the use of
search engines on one hand and using artificial intelligence on the other. ‘A’
is using Google for the research and ‘B’ is using ChatGPT (a commonly used
artificial intelligence) for academic research. ‘A’ has used Google and has
been redirected to a page from where the content has been copied and ‘B’ as
well has copied the content from ChatGPT. Both ‘A’ and ‘B’ have submitted their
work as research to their institution. The content has been subjected to
plagiarism check and ‘A’ has got report that the content is 50% similar to the
content in the website ‘Y’ and ‘B’ has got report that the content is 50% AI
generated.
So, what difference does it mean and what are the consequences of
research? ‘A’ though has used Google for research has copied content from
website ‘Y’ which Google suggested and the similarity check doesn’t recognise
Google as the content provider but ‘B’ has used ChatGPT and the report shows as
AI generated. Therefore, the major difference between search engine and
Artificial intelligence is the recognition given to artificial intelligence as
a primary source of content, unlike
Google which is not recognised as the primary source of content provider.
This makes anyone think about why it is like that and what is the
status of artificial intelligence and whether it has protection and liability
over generated content. Before understanding the authorship and ownership of
the content generated by artificial intelligence, it is important to understand
the need of why authorship and ownership need to be given to the artificial
intelligence generated content.
In illustration[6] 1 @
footnote 5, when ‘A’ has used website ‘Y’ and later was found that the content
was incorrect and was brought to the notice of the website officials and can be
removed/altered as required and the officials can apologize to the concerned
people and public based on the impact of the content posted in the website.
‘B’ has generated the content and later was found offensive to the
public interest. What next? Information was not posted by anyone until the
input for generating such content was given. So, is there no remedy for such an
error? Who is to be held liable and who needs to remove such content and either
compensate and apologize to the public?
When artificial intelligence creates wonderful content, which is
appreciated by the public and gets recognition, both the company owning the
artificial intelligence and the user who prompted artificial intelligence for
such content creation would want to claim authorship and the monetary benefits.
But when the content is irrelevant and is causing public outrage who is
responsible? Either authorship and liability someone needs to take it and there
must be clarity in this regard as people in the rapid growth of technology and
artificial intelligence must not be confused and afraid of using the
technology.
3.
AUTHORSHIP CONCERNS
Artificial intelligence has got no proper recognition as a
separate entity in any jurisdiction as a human or artificial entity. In Berne
Convention[7], it refers
to a natural person as an author of creative work. In the case of ‘EBC v.
D.B. Modak’, the court laid down that there must be a minimum “Modicum of
creativity” along with innovation. As definition of artificial intelligence at
footnote 2 specifies that artificial
intelligence uses the existing data and interprets it to provide the required
output as defined in the footnote 3. In
Section 2(d)(vi)[8] of Indian
Copyright Act, 1957 (the Act) it mentions who can be an author. Section 17[9] of the Act
defines an owner.
What is artificial intelligence and to what can it be compared for
getting the legal status of an entity in the scope of existing legislations. We
can think of the company as an artificial person in the Companies Act when we
think of something artificial entity. Therefore, let’s compare artificial
intelligence with that of a company to identify if artificial intelligence can
be identified as a company or not to get legal status.
To better understand, let's illustrate[10] it with an example. Section
2(20)[11] of the
Companies Act, 2013 defines the company. Let ‘C’ be a company registered under
the Companies Act, 2013 and ‘D’ be an artificial intelligence to be registered
as a company. Here, we will compare both ‘C’ and ‘D’ with the salient features[12] of a
company and see if both have same features of a company or differ somewhere.
- SEPARATE LEGAL
ENTITY
‘C’ a company registered under the Companies Act, 2013 has a separate legal entity from its members. ‘D’ an
artificial intelligence can be considered to have separate legal entity as it
can function on its own name and exist independently.
- LIMITED LIABILITY
‘C’ has limited liability as it is a registered company. ‘D’ being
an artificial intelligence that can only
function with human input and database, can be said to have a limited liability as without the database and input it
cannot function.
- PERPETUAL
EXISTENCE
‘C’ has perpetual existence until the winding up of the company.
‘D’ being an artificial intelligence can be said to have perpetual existence
until it is completely deleted.
- SEPARATE PROPERTY
‘C’ has a separate legal entity and therefore, can acquire
property separately. ‘D’ being able to be a separate entity can acquire
property for maintaining database and servers.
- SHARES
‘C’ being a company has shares and can be transferrable. ‘D’ being
able to be recognised as a company, can have shares and be transferred.
- CAPACITY TO SUE
AND BE SUED
‘C’ has the ability to sue and be sued as a company and ‘D’ too
has the ability to sue or be sued as it has no difference to that of a company.
- COMMON SEAL
‘C’ has a common seal as it is a company and ‘D’ being able to be
recognised as a company can have a common seal and as it is particularly having
online existence can have its digital signature.
As we have compared artificial intelligence to a company and have
found positive results, we can consider artificial intelligence to be a company
in the scope of existing legislations. Further, we need to analyse whether an
artificial intelligence registered as a company can have authorship and
ownership over the content generated by it.
Let ‘D’ referred to in illustration 2 at footnote 10 be considered as a service-oriented
company. A service-oriented
either has online or offline presence, provides its clients/customers with
services and no physical product. ‘D’ being an artificial intelligence has
online presence and provides content to the
users, it can create content which are copyrightable.
Other services cannot be copyrighted but literature, art and music
are under the scope of copyrightable expressions. Many similarity and
plagiarism checkers detect the content generated by artificial intelligence as
a separate category as “AI generated”. If the tools don’t recognise the content
as “AI generated”, does it solve the problem of the authorship of artificial
intelligence generated content? This will be discussed later.
Further whether ‘D’ can get copyright for the content or not is a
question though the content is copyrightable under the Act. For this, we need
to identify why the content
created by humans is copyrighted and whether the artificial intelligence can fulfil
such criteria. This can be understood with the ratio decidendi of the cases and
the rationale behind the copyright.
4.
RATIONALE BEHIND GRANTING COPYRIGHT[13]
Copyright granted to any individuals grants them with economic
rights u/s 14 of the Act and moral rights u/s 57 of the Act. Copyright
protection enables the author to protect their creative expression and
encourages innovation and creativity. Copyright protection acts as a reward for
the creative work and enables authors to benefit from economic exploitation and
be recognised for their work. In the case of ‘RG Anand v Delux Films & Ors.’[14], it was held that only
expression of an idea can be copyrighted and not mere idea.
Principles of originality and the rationale of copyright
Sweat of the Brow Test
In ‘University London Press
v. University Tutorial Press’[15], it was held that copyright can be granted for
the labour put into the work[16]. This was
followed even in the Indian courts[17].
Modicum of creativity
In ‘Feist Publications Inc.
v Rural Telephone Services Co.’[18], it held that there must be
intellectual effort and degree of creativity.
Skill and Judgement Test
The author must have applied his ‘skill and judgement’ in creating
the work and the work must have a degree of creativity.
Applying these principles to the work created by humans and
artificial intelligence, it can be understood that human applies labour,
intellectual effort, skill and
judgement with a degree of creativity unlike artificial intelligence which
collects data from various sources and regenerates the same information.
To understand this, let us illustrate this with an example, ‘D’ an
artificial intelligence-oriented service provider and ‘E’ an individual
researcher. Both ‘D’ and ‘E’ are subjected to a test to find an answer to a
question for which there is no direct answer and one needs to refer to various
materials and interpret them to come to a rationale conclusion.
‘D’ with the data input in the server interprets it and provides a
result (‘F’) and ‘E’ refers to the online materials along with offline
resources and with human understanding and efforts provides a result (‘G’).
Here, the question of correctness of the result arises along with what
information has been used.
‘D’ has only used the data that has been stored in the database
and interprets them to provide a result. And what if the data is wrong o there
is no data stored. It shows a wrong result based on the data stored in the
database or shows error and wouldn’t be able to understand and interpret the
question. This makes it clear that artificial intelligence functions only when
the data is stored and can only interpret them and nothing beyond that.
‘E’ has used the online and offline resources and has provided a
result with human understanding. What if the information accessed is wrong? ‘E’
can interpret the question and the information and with human understanding can
rectify the information and provide a valid answer or raise objection about the
question.
One might think that if the information is correct and there is
data stored in the database, then there is no problem and artificial
intelligence can beat human intelligence and ease the work of a human.
When ‘D’ and ‘E’ are provided with all the information and then
subjected to the same test, is the result same? No, when both are subjected to
an unethical and illegal question, ‘D’ still provides a result from the
database and ‘E’ with human intellect interprets the question and then decides
what the response must be and whether to respond or not. Therefore, though
artificial intelligence is provided with all the information there needs to be
a human intellect behind while storing the information in the database and when
using such information after the result is generated through artificial
intelligence.
This makes it clear that artificial intelligence doesn’t qualify
the principles of originality as there is no creativity, skill,
judgement nor labour. So, the content generated by
artificial intelligence doesn’t qualify for copyright?
There’s more to it. What about the economic rights and moral
rights of the work? Though there is less scope or no scope for artificial
intelligence to be granted copyright, the rights for the work must be granted to someone. Who is it? The artificial
intelligence or the company maintaining the artificial intelligence or the
person giving the prompt or the government?
ARTIFICIAL INTELLIGENCE
The the information is created based on the information stored in
the database and the input given. But does it mean it isn’t creating any work
and is not eligible for copyright and the rights conferred in it?
Let us illustrate[19] an
example for better understanding. Let ‘D’ be an artificial intelligence tool
and ‘G’ and ‘H’ are two individuals who are generating responses from ‘D’. The
database stored in ‘D’ is same for ‘G’ and ‘H’. Based on the similar inputs
given by ‘G’ and ‘H’, will the results be same?
No. Though the input is similar, the artificial intelligence
learns from the user interaction and the information generated is unique and is
based on the user and the interaction. So, should artificial intelligence get
rights conferred in copyright? No, as artificial intelligence generated content
cannot fulfil the principles of originality, it cannot be granted copyright
thus no rights.
What if it is conferred with rights? Every output generated by
artificial intelligence needs to be granted copyright making it have millions
of copyrights on a daily basis for every single use. So, there is a concern
about
granting copyright to artificial intelligence and
therefore no rights can be granted and artificial intelligence cannot exploit
economic rights and moral rights independently and what other issues are faced
if granted with copyright.
What is the duration of copyright? TRIPS has set minimum standards
of copyright protection as a lifetime of the author plus fifty years and
artificial intelligence having a perpetual existence, cannot be granted
copyright protection forever as considering artificial intelligence as an
author, the information must be protected and cannot be made public.
Though artificial intelligence is not granted copyright protection
and economic rights, it is being granted moral rights when content generated by
artificial intelligence is being recognised as AI generated. But the issue is
that it doesn’t specifically recognise which artificial intelligence is being
used. So, whether it can be considered as getting moral rights or not?
This issue can only be solved when either the artificial
intelligence is recognised so every artificial intelligence is having different
identity and can be identified separately or each artificial intelligence have
different and unique way of response compared to other artificial intelligence
so that the tool detecting the content as AI generated can specifically recognise
which artificial intelligence is generating content.
Thus, granting authorship to the artificial intelligence has many
issues and concerns but who takes the liability for misinformation and other
errors? This will be discussed later.
THE COMPANY
The company developing the artificial intelligence has developed
codes for the artificial intelligence and has stored information in the
database and given the specifications and other automated prompts for
generating response in no time. The company has made all the arrangements for
the artificial intelligence to operate independently without regular
monitoring.
Does this mean the company doesn’t have any liability or need to
regulate the artificial intelligence? First to understand the liability we need
to know about the authorship.
To understand the authorship, let us illustrate an example. Let
‘I’ be a software developing company and ‘D’ is an artificial intelligence
developed by ‘I’ and ‘J’ is a normal gaming software developed by ‘I’. Both are
being launched to the public.
Before a software is launched, the company tries to identify
errors and does small scale testing and make sure that the software doesn’t
crash, malfunction or create any error in any system. So, there is a reasonable
care and checks that there are no errors. And there is always an update and
developments in a software. So, this makes the company eligible for the
authorship over software.
But that is over a software and not the way of how the software
functions. This makes it clear that the company has authorship over artificial
intelligence but what about the content? The company through the software i.e., artificial
intelligence provide service to the public and the company has all the rights
over any service provided. This means that though the public gives input to the
artificial intelligence to generate responses, it can be understood that
providing inputs doesn’t involve creativity, skill and judgement and intellect
effort as mentioned in the principles of originality and the company for
developing the artificial intelligence and putting efforts, creativity and
skill and judgement in giving right instructions and ways to respond, the
company is capable of being granted authorship of the content.
But the artificial intelligence is a service and is considered not
eligible to get copyright. So, irrespective of the efforts over the development
of artificial intelligence the company cannot claim copyright over the content.
So, what does the company get in return for the efforts put into the development
of artificial intelligence?
The company has authorship over the artificial intelligence in the
category of software in the copyright legislations. So, the company can exploit
the software and the codes of the artificial intelligence and acquire economic
rights and moral rights.
Thus, the company doesn’t get copyright over content generated by
artificial intelligence but can it take liability for the error in the
responses? This will be discussed later.
INDIVIDUALS
As discussed earlier, giving any prompt doesn’t involve
creativity, but what about skill, judgement and intellect efforts?
Getting response from an artificial intelligence is a difficult
task. How many people are techies or into electronic gadgets and can make full
use of them. This makes anyone think about whether they are capable to use technology to the fullest
or not. When two individuals are given same technology and same problem, there
is a difference in the results and even the time to get a proper response
matters as that determine the skill and judgement and intellect efforts of
humans.
When two individuals are using the artificial intelligence, one
might get the right answer in no time and the other might get any error or
wrong response. This can be a determining factor of having intellect efforts
and skill and judgement and to an extent creativity as giving a correct prompt
is a creative input of human.
So, this further leaves us with whether individuals be granted
copyright for the content. This is a complex area of understanding as everyone
who uses artificial intelligence will be generating responses and everyone
needs to get copyright for the content they generate and this piles up the
copyright claims and this reduces the value for the principles of originality.
As artificial intelligence is a free to use software, the use of
the software must be for personal use and it must be upon the individual to
make best use of it. So, what if artificial intelligence still provides wrong
information, who should be liable?
There is an extent of development of artificial intelligence where
the software has premium subscription services as well. So, in a free version
any individual who uses artificial intelligence has to own the liability or use
it to only personal use or have restricted use? Is there any difference between the free and premium version? There is no difference in
he context we are discussing. Therefore, both free
and premium versions are considered same. So, if anyone uses premium version
and gets wrong response will the artificial intelligence or the company be
liable?
No. The terms and conditions of the artificial intelligence for
both free and premium versions expressly say that “ChatGPT can make mistakes. Check important info.” Therefore, both
the company and the artificial intelligence are not liable as it is the
discretion of the user to rely on the information. So, does it make the user
liable for the content though not being eligible for authorship?
GOVERNMENT
What does government have the copyright over the artificial intelligence
generated content mean? It means that the government can allow the work to be
published without the consent of the author through compulsory licensing.
Compulsory licensing is a provision in copyright law that
government grants a licensee to use a work of copyright without requiring
approval from the owner of the copyright. Section 31[20] of the Act states when can
compulsory licence be granted. As there is no author for AI generated content,
can it be considered that it can be compulsorily licensed by the government or
and agency?
In Section 31A[21] of the
Act defines for compulsory licencing of unpublished or published work when the
owner cannot be traced, dead or not found, etc. As there is no proper author
for the artificial intelligence generated content, the content can
automatically be granted compulsory licence u/s 31 and 31A of the Act and be
made public.
But this further leads to the question of quality and validity of
the content generated by artificial intelligence. So, if the content generated
by artificial intelligence is to be granted compulsory licencing, it means that
the content is considered correct, creative and has qualified the principles of
originality. Therefore, artificial intelligence be given status of author. Who
monitors the content and does all the content generated be granted compulsory
licensing? Artificial intelligence being an online platform and having
international presence has millions of users using the tool and there will be
billions of responses generated on a daily basis. Does every response get
compulsory licensing and be made public or there needs to be scrutiny? Is it
possible for scrutiny of billions of responses on a daily basis?
No. Only exceptional responses must be getting compulsory licence
and others must be treated as non-creative content and must be made public or
disregarded? If disregarded there is no problem. But what difference does it
make when both exceptional content and basic content are made public? The
issues of rights and liability arises again.
There is always a lack of clarity regarding authorship and
liability of artificial intelligence generated content. We have tried to
identify the authorship for the artificial generated content. We need to
understand the authorship from the perspective of academic integrity because
academic and research is always relying on the content and it is important to
understand the value of artificial intelligence and how reliable and who needs
to be credited for the content from artificial intelligence.
As we have already discussed as to who can be author among
artificial intelligence, company maintaining the artificial intelligence or the
government. We can further the discussion on the same grounds.
ARTIFICIAL INTELLIGENCE
As we have discussed the problem faced in considering the
artificial intelligence as the author but it is being recognised as the source
in the similarity and plagiarism checker making it an author. So, can
artificial intelligence be considered as author and be credited for the
information. Does it mean that giving due credits to the content as AI
generated make it not plagiarised?
Though the tools recognise the content, it doesn’t specify which
tool in specific has generated the content and how to give credits to the
artificial intelligence?
It generates content based on the database and the input and can
artificial intelligence be considered as author and the tool as blog and the
chat page as the journal name or is it sufficient to just post the link? So, if
artificial intelligence is being recognised as a source, it must be considered
as author and be given credits. But this again leads us to the question of
copyright and liability concerns.
So, why recognise the content as AI generated when it can’t be
considered as author and be given credits? The answer is to recognise as
generated by artificial intelligence so the concerned authorities need to cross
check the information and not completely rely on the information. And is it to
not recognise the human as the author?
Though there are assumptions regarding the recognition given to
artificial intelligence there is still no clarity as to why artificial
intelligence is being recognised when it can’t be considered as author.
THE COMPANY
As we have seen whether the company can be considered as author or
not, it is clear that the tools recognise as AI generated content and not specifically
recognise which company does the artificial intelligence belong to. Therefore,
the company can’t be given credits?
When a judge writes a judgement, it will be in the names of the
judges and not in the name of the court or judiciary. So, on a similar ground,
the company cannot claim the authorship and credits?
The company has it is being given rights for the artificial intelligence and
can exploit the software it can’t exploit the responses considering it as a
service provided to public.
INDIVIDUALS
As we have seen how individuals cannot be granted authorship for
the content, it must be understood in terms of credits. The content is being
recognised as AI generated and this reduces the reputation of the individuals
using that content. But we discussed as to how generating a proper response
require skill and judgement and intellect efforts.
But we need to understand that research happens based on the
existing information unless it is empirical and research on any new area. So,
any research needs analysis of the existing data and rebuilding new data and
content with different perspective or criteria. So, the individuals who are able to use the artificial intelligence to their best to
generate responses can be considered as researchers and can be granted credits.
But does having smartness on using technology and artificial
intelligence make an individual a researcher like an individual who puts their
efforts, time, creativity and does field research? There is lots of difference
in an individual generating responses on their fingertips at the comfort of
their homes and an individual who does field study, collect data etc.
Therefore, as individuals cannot be granted authorship, they cannot be given
credits for the work they have not done.
And the other problem in giving credits to individuals is that the
chat with the artificial intelligence is private and only whether the content
is by artificial intelligence or not can be identified based on the response
pattern. To recognise individual as an author or to give credits, there comes
privacy concerns as the chat with the artificial intelligence needs to be
monitored and there can be multiple people generating same content and who
needs to get credits is a question.
THE GOVERNMENT
As discussed, it doesn’t mean the government becomes the author
and get credits for the work. The government through granting compulsory
licensing make the work public and available for public use.
So, should the government get the credits and authorship to the
work? No. The government through an authorised agent or publisher publishes the
work and make it public and it becomes available to public use. The money
generated by publishing such work can be claimed by anyone by proving that they
or their known person have created he work.
But when a work is being made public and free to use or available
to public at cost by government recognised agent the work is considered as
valid and correct. So, what if the content is wrong? It needs to be taken down
again and declared incorrect. Until the content is removed, the content is
considered as correct and copying that content must be considered as valid and
not be recognised as plagiarism by any tools.
So, does this make any content generated by artificial
intelligence be made public by government through compulsory licensing be
considered valid so making anyone using artificial intelligence for their
research make them not liable for using artificial intelligence and it means
that the individuals be credited for reviewing and commenting on such content.
Therefore, even after a detailed study on the authorship of the
content generated by artificial intelligence is an area of further study and
there needs to be clarity given by any appropriate authorities and there needs
to have necessary changes in the legislations to incorporate the decision given
by authorities.
Though there is a lack of clarity on authorship and credit for the
content generated by artificial intelligence, there must be some clarity on
liability of the content generated. The study will further focus on the
liability aspect of the content generated by artificial intelligence in the
same pattern as discussed for authorship and credits.
ARTIFICIAL INTELLIGENCE
This is an independent tool capable of performing on its own without
regular monitoring by any human. Therefore, can be said to have a separate
identity. But this doesn’t make it have authorship over its content. Does this
exempt it from liability of error in response?
If it doesn’t even have a separate identity in the eyes of the law
and is not given rights, how can it have liability which is a basic principle
of jurisprudence. So, whatever the content is generated, if any error occurs
who is liable? Even though the artificial intelligence is liable, how can a software
be made liable for the errors in the codes and mechanism of its functioning and it is indirectly the human behind the
software who can be liable.
THE COMPANY
As we saw that the software cannot be liable and only human can be
held liable, it must be developer and tester who is liable and has it is in the
course of employment, it must be the company who must be vicariously liable
along with the developers and testers for the error in the software.
But the company has clearly mentioned that the software can make
mistakes and user need to check information and the company doesn’t regularly
check or monitor every chat and response, so the company cannot be held liable.
But when any input is generating wrong response, the company needs to be
informed and there must be corrections made and if they don’t make correction,
then they can be held liable based on the safe harbour principle.
INDIVIDUAL
Though the software mentions that the user needs to check the
information, people use the software for easing the work and if they have to
check every information then they can use other resources available in the
library and online resources. So, using he software doesn’t make the user
liable?
When an individual uses the software for personal understanding of
any topic or other personal use, then there is no problem. But using the
content for commercial use and publishing it as own content without checking
the content for validity or existing copyright or other issues concerning the
content the individual must be held liable to an extent of the offence.
GOVERNMENT
Though the rule of law principle says that no man is above law,
the government is just facilitating the reach of the content to the public
through compulsory licensing and therefore, cannot be held liable for the
content which is not created, monitored and have monetary benefit to the
government.
So, the question of liability is clear as compared to that of
authorship and credit for the content generated by artificial intelligence. As
we have seen authorship, credit and liability of the content, we need to study
the copyright concerns and need of the copyrighted content for the smooth
functioning of artificial intelligence.
We have understood that the artificial intelligence functions
mainly based on the content stored and interpreting such content as per the
requests given by user. So, it can be clear to anyone how important is the
content for generating any responses. One might think that artificial
intelligence is just like any search engine but search engine just provides
relevant sources as results from which the user needs to refer and take content
based on the availability and restrictions of those sources.
To understand the need of content and the copyright concerns we
need to understand the difference on how the artificial intelligence and search
engine work. Search engine when posed a question, it redirects the user to page
having various sites and the sites has information. What does it mean?
When a user searches for any information which is general and freely
available, the search engine shows various sites from which the user can refer
for information. And the search engine is not liable for the content in those
sites and it just acts a bridging agent between the user and the sites. What if
the content is paid and not available?
The user needs to pay for the content in the particular site if
they wish to buy the content and there are many sites which provide access to
the users for reading but has restrictions in copying the content and has to
pay and buy the article for using such content. So, there are many ways of
restricting the use of content and there is necessity of buying the content for
using the content.
In search engine, the user has the option to choose the content
and sites from where the content can be taken either free or paid and in
artificial intelligence the user doesn’t have the option to choose the sites
and type of content the artificial intelligence uses and it generates making it
crucial to have access to various content and many contents might have copyright
and be having restricted access.
But the artificial intelligence needs to be having access to such
copyrighted content to be providing real time response and generate unique and
correct responses. Unable to get the access to copyrighted and premium content,
there are high possibilities of wrong responses. So, what about the monetary
rights of the copyright holder?
The company developing the software must take licence from the
copyright holder and charge minimal charge from the users or the use of such
content must be exempted from infringement considering the use as research and
non-commercial use under the fair use and fair dealing principles as
individuals can be held liable as discussed earlier. There must be a way to
balance the economic rights of the copyright holder and the public interest in
the developing era of technology or else, the best use of artificial
intelligence can never be made and it just becomes a search engine or site like
Wikipedia where anyone can write and edit content and the content has no value
for academic and research and can be used only for basic understanding of any
topic.
5. CONCLUSION
This article through a detailed study has focused on the
authorship, who needs to get credits for the content generated by artificial
intelligence and who takes liability for the mistakes in the content and to
what extent. This article has considered the detailed study on the artificial
intelligence, the company maintaining the artificial intelligence, the
individual using the artificial intelligence and what role does government play
in regulating the content and balancing public interest in terms of authorship
and liability of the content generated by artificial intelligence. This article
focuses even on the copyright concerns of the content generated by the
artificial intelligence and the content used by the artificial intelligence and
the need to balance economic rights of the copyright owner of the content used
by artificial intelligence and the public interest in the development of
technology.
This article through this study recommends that the concerned
authorities consider the need of bringing clarity about the authorship and
status of artificial intelligence and the liability for the content generated
by artificial intelligence and to what extent the user can rely on artificial
intelligence and what is the status of infringement concerns of using
copyrighted content in the database of the artificial intelligence. As the
technology is developing, it is important for the users to understand the good
and bad side of anything and their responsibilities and safety measures.
[1]
White Paper on the use of Artificial Intelligence in Trade Facilitation (2024)
[2] “Artificial intelligence (AI) systems are
software systems designed by humans that, given a complex goal, act in the
physical or digital dimension by perceiving their environment through data
acquisition, interpreting the collected structured or unstructured data,
reasoning on the knowledge or processing the information derived from this
data, and deciding the best action(s) to take to achieve the given goal. AI
systems can either use symbolic rules or learn a numeric model, and they can
also adapt their behaviour by analysing how the environment is affected by
their previous actions.”
[3] “Discipline of computer science that is aimed
at developing machines and systems that can carry out tasks considered to
require human intelligence, with limited or no human intervention.”
[4]
Trends and Developments in Artificial Intelligence – Challenges to the
Intellectual Property Rights Framework (2020).
[5] For
reference in the future, will be referred as illustration 1 @ footnote 5.
[6] For
future reference, will be referred as Illustration 1 @ footnote 6.
[8] “In relation to any literary, dramatic,
musical or artistic work which is computer-generated, the person who causes the
work to be created”
[9] Subject to the provisions of this
Act, the author of a work shall be the first owner of the copyright therein:
Provided
that--
(a) in the case of a
literary, dramatic or artistic work made by the author in the course of his
employment by the proprietor of a newspaper, magazine or similar periodical
under a contract of service or apprenticeship, for the purpose of publication
in a newspaper, magazine or similar periodical, the said proprietor shall, in
the absence of any agreement to the contrary, be the first owner of the
copyright in the work in so far as the copyright relates to the publication of
the work in any newspaper, magazine or similar periodical, or to the
reproduction of the work for the purpose of its being so published, but in all
other respects the author shall be the first owner of the copyright in the
work;
(c) in the case of a work made in the course of the authors
employment under a contract of service or apprenticeship, to which clause (a)
or clause (b) does not apply, the employer shall, in the absence of any
agreement to the contrary, be the first owner of the copyright therein;
[10] For
reference in the future, will be referred as illustration 2 @ footnote 10.
[11]
“Company means a company incorporated under this Act or under any previous
company law”
[13] The
principles are referred from Robbin Singh, “UNDERSTANDING
THE CONCEPT OF ORIGINALITY UNDER COPYRIGHT LAW IN INDIA”, Law Mantra think
beyond others available at *11.pdf (lawmantra.co.in).
[14]
R.G. Anand v. Delux Films & Others, AIR 1978 SC 1613
[15]
University London Press v. University Tutorial Press, [1916] 2 Ch 601.
[16]
“Merely because time, energy, skill and labour were expended (ie, originality
of skill and labour).”
[17]
Burlington Home Shopping v Rajnish Chibber. 61 (1995) DLT 6.
[18]
Feist Publications Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991)
Feist Publications Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991).
[19] For
future reference, will be referred as illustration 3 @ footnote 19.
[20] “Compulsory
licence in works withheld from public. -
(1) If at any time during the term of
copyright in 1[any work] which has been published or performed
in public, a complaint is made to the 2[Commercial Court] that
the owner of copyright in the work--
(a) has refused to republish or allow
the republication of the work or has refused to allow the performance in public
of the work, and by reason of such refusal the work is withheld from the
public; or
(b) has refused to allow communication
to the public by 3[broadcast] of such work or in the case of
a 4[sound recording] the work recorded in such 4[sound
recording], on terms which the complainant considers reasonable,
the 2[Commercial
Court], after giving to the owner of the copyright in the work a reasonable
opportunity of being heard and after holding such inquiry as it may deem
necessary, may, if it is satisfied that the grounds for such refusal are not
reasonable, direct the Registrar of Copyrights to grant to the complainant a
licence to republish the work, perform the work in public or communicate the
work to the public by 4[broadcast], as the case may be, subject
to payment to the owner of the copyright of such compensation and subject to
such other terms and conditions as the 2[Commercial Court] may
determine; and thereupon the Registrar of Copyrights shall grant the 5[licence
to such person or persons who, in the opinion of the 2[Commercial
Court], is or are qualified to do so] in accordance with the directions of
the 2[Commercial Court], on payment of such fee as may be
prescribed.”
[21]
“31A. “Compulsory licence in unpublished [or published works]. — [(1) Where, in the case of any
unpublished work or any work published or communicated to the public and the
work is withheld from the public in India, the author is dead or unknown or
cannot be traced, or the owner of the copyright in such work cannot be found,
any person may apply to the [Appellate Board] for a licence to publish or
communicate to the public such work or a translation thereof in any language.]”
