Open Access Research Article
JUSTICE IN AI- GENERATED WORLD AND ITS LEGAL LANDSCAPES BY: SIDDHI JAISWAL
Published Paper
Article Details
JUSTICE IN
AI- GENERATED WORLD AND ITS LEGAL LANDSCAPES
AUTHORED
BY: SIDDHI JAISWAL
Law
Student
Narsee Monjee
Institute Of Management Studies, KPMSOL, MUMBAI
Abstract
The rapid
advancement of Artificial Intelligence (AI) has ushered in a new era of
creativity, where machines can generate content that rivals human-produced
works. This technological revolution presents a complex legal challenge such as
the difficulty in distinguishing between AI-generated evidences and non AI
generated evidences. Since the AI-powered tools have become more sophisticated,
they can produce deepfake videos, images, audios and other forms of fabricated
content that can be easily mistaken for authentic materials and be presented as
evidence in the courts. This raises significant questions about the credibility
and admissibility of AI-generated evidence in legal proceedings. Moreover, the
paper delves into the emergence of AI-generated content challenges traditional
notions of authorship, originality, and ownership. AI systems can generate
creative works, such as text, images, and music, that may qualify for copyright
protection under IPR. However, the lack of human authorship raises complex legal
questions about who owns the rights to such works and who must be held liable
in case a civil or criminal offense has been committed which is discussed
through case studies by the author. Furthermore, this paper aims to contribute
to the ongoing discourse on the AI generated evidences and how it would
challenge the substantive IP laws in picture in the digital age. The paper also provides recommendations for
policymakers, legal practitioners, and AI developers to navigate the
complexities of AI-generated content.
Keywords
- Artificial Intelligence, evidences, admissibility, ownership, copyright, IPR.
The
urge for legal reform following the rise of AI systems as creators and
contributors across different spheres of life has severely challenged the
traditional legal framework. It covered two major areas on these challenges:
admissibility and probative weight of AI-generated evidence in courts and
disruption of intellectual property laws by AI-generated content. Thus, a
combination of cases and analyses demonstrates how AI deepfakes, fabricating
evidence, and art raise more nuanced questions with respect to the principles
of authenticity, authorship, ownership, and liability.
With respect to legal evidence, it is the very
sophistication of AI technologies which in turn complicates the courts’ task of endorsing authenticity by distinguishing
between what is genuine and what has been fabricated. The already-existing laws
primarily create hindrances for the identification of AI-generated evidence considering
originally when developed for those by human beings. So do the laws of IP,
derived from individual humanistic concepts of authorship and originality,
conflict with the dimension of AI being more contributory with less human
involvement than was sanctioned before. Such a gap, anachronistic regression
from new methodologies, is the prime manifesto for much-needed reformation. To
deal with these problems, a balanced approach using legal, technological, and
ethical solutions is imperative. The courts must adopt advanced mechanisms and
expertise with respect to the evaluation of AI-generated evidence;
restructuring of IP laws is necessary to make room for the much-invoked AI
systems which are contributing too much to the creative process. Stakeholders-including
policymakers, practitioners of law, and AI developers-need to work together to
ensure that the guidelines are unambiguous, aligned, and equitable for all
other parties concerned. However, adjusting the law is a question for a new
evolution to be conceived for handling AI content, preserving the integrity of
courts and intellectual property control. Progressively approaching such reform
can enable AI to unleash its own revolution without any compromise on the
standards of justice, equity, and creativity.
Thus, an exploration of the interaction between AI, legal
frameworks, and social expectations tends to these concerns in all their
multifaceted dimensions. By closely examining such actual situations as those
of DABUS and disputes concerning deepfake evidence, it critically evaluates
these quandaries presented by AI content integrating within both the judicial
process and intellectual property law. The analysis illustrates how AI
questions traditional legal concepts, points to inconsistencies across
jurisdictions, and suggests the need for urgent reforms. In summary, the aim of
this study, therefore, is to serve to clarify the botched and tortured
relationship between AI technology and the law, with designed recommendations
for policymakers, legal professionals, and AI developers. This paper will
significantly inform the conversation about how to regulate AI technologies in
a morally, socially, and legally acceptable manner by addressing issues of
AI-generated evidence credibility, authorship and ownership of AI-created
works, and accountability of responsible parties.
1. MAURA R. GROSSMAN (2023) The paper
discusses the evidentiary issues that must be addressed by the bench and bar to
determine whether actual or asserted (i.e., deepfake) GenAI output should be
admitted as evidence in civil and criminal trials and offers practical
recommendations for courts and attorneys in meeting the evidentiary challenges
posed by GenAI.
2. Akerman LLP (2024) This article
delves into the multifaceted challenges presented by AI-generated evidence,
explores the principles governing its admissibility. It suggests the
development of clear standards and regulations, educating legal professionals
on the capabilities and limitations of AI and promoting the development of AI
in an ethical manner.
3. V.
K. Ahuja (2020) The article discusses the legal position in several countries
and deals comprehensively with several models of authorship in AI-generated
works. The discussions being made in this regard at the WIPO have also been
discussed in the article.
4. Paul
W. Grimm (2021) This paper explores issues that govern the admissibility of
Artificial Intelligence (“AI”) applications in civil and criminal cases. It
provides a detailed yet intelligible discussion of what AI is and how it works,
a history of its development, and a description of the wide variety of
functions that it is designed to accomplish.
5. Eftychia
Bampasika (2020) This paper served the purpose of highlighting the interplay
between the criminal justice systems and the AI technology in connection with
AI being employed as evidence tool which can pose a threat to the justice
system.
The author has used secondary research and has collected
information through case studies, articles, and research papers by other authors. The author has
read and analyzed these sources and then formed an analysis with respect to the
particular topic for research which will serve as primary data. The research is
qualitative and comparative in nature. The citation format is 4th Edition
OSCOLA.
·
To explore the admissibility and credibility
AI-generated evidence in legal proceedings.
·
To determine the evidentiary principles required
for addressing admissibility of AI evidences.
·
Addressing the challenges faced by AI evidence
in court as admissible.
·
To investigate the impact of AI-generated works
on traditional notions of authorship, originality, and ownership under
intellectual property laws.
AI-generated
content presents substantial challenges to legal systems, particularly in terms
of the authentication and admissibility of evidence. Additionally, it disrupts
conventional intellectual property frameworks by challenging the established
notions of authorship and ownership.
The state or characteristic
of being acceptable or admissible is referred to as admissibility. From the
lawman's definition, "evidence" is anything the court accepts to
affirm or deny the alleged facts in a trial. Consequently, any evidence in the
form of a document, or physical appearance presented before a Court of Law can
be termed admissible evidence. In fact, the Court of Law allows only relevant
and credible evidence in the courtroom; all other evidence is not allowed. The
BSA defines the legal term "admission." The term refers to the
authority or permission to approach, enter, allow, or enter. Legally speaking,
agreement or concurrence with another person's statement is different from a
confession in that the former requires previous investigation by another party,
whereas the latter does not. An admission made outside of court that is
accepted as evidence can be a fact, point, or statement. Electronic evidence
has been a crucial component of legal procedures since the invention of
electronic devices and the internet. Electronic mail, SMS, security film, and
even social media posts, which are mostly utilized to ascertain facts and make
legal arguments in the court of law, contribute to the cause of this particular
challenge. Indeed, the emerging AI-generated evidence presents a whole new
challenge when electronic evidence grows increasingly popular.
Electronic evidence becomes a
fundamental element of modern-day cases in court. Demonstrative facts are
combined with an argument or case by using emails, texts, social media posts,
and even security footage. While the judicial system has procedures in place to
manage electronic evidence, there is now a new level of complexity added to that
through AI-generated evidence. The growing complexity of AI-generated content
also poses serious problems for the legal system, particularly in the
determination of reliability and admissibility. Traditionally, courts relied on
the human knowledge and well-established forensic instruments. However, the
courts are now faced with the task of distinguishing between real and
artificial intelligence-generated content. A capable litigator, before stating any matter would see how the facts
could be presented as a good evidence. Although the parties give the lawyer
appropriate documents and tell their story, the rules of evidence may in
certain cases demand much more. More often than not, additional witnesses are
required to avoid hearsay objections or to authenticate texts or written
documents. With the advent of expert testimony, the considerations become
increasingly complex. AI raises multiple questions about the legitimacy of the
evidence that it might present in court. AI's inferences will not be acceptable
if it is not established that it's reliable. For instance, if AI is to be taken
to understand any given social media posts, it could be adjusted in such a way
that it searches for certain keywords or phrases referring to particular
behaviors.
JUDGES PERSPECTIVE
It
is difficult for judges to decide whether such evidence can be admitted in
court due to problems of bias, interpretability, transparency, and
dependability. Generative AI has also increased false information and incorrect
data, making it hard for judges to believe that it is true. Recently,
generative AI created deepfake, sexually graphic photos of singer and
songwriter Taylor Swift, which was reported in the media. A picture of Pope
Francis in a clean white jacket caused a buzz since it appeared to be an
authentic photograph.
Imagine what if the picture depicts a public figure
committing a crime or any illegal activity? How might the court or lawyer
demonstrate that this picture is authentic when this is the case? If this is an
AI-generated image, how will the judge know? In addition, the opacity of AI
algorithms makes it impossible to be transparent, and bias in training data may
lead to discriminatory findings besides the many dangers that compromise the
validity and dependability of evidence. There are no established protocols for
confirming AI-generated evidence, which makes decision-making difficult.
Another very practical example of the issues with electronic evidence is
self-driving automobiles. For instance, it remains unclear how results from a drowsiness
detector will be applied in adversarial or inquisitorial legal systems to
determine liability for an accident. We have to check the data of the AI system
for correctness and limits, assign blame for mishaps or disagreements, and
understand the logic behind its judgments.
SIGNIFICANT
ISSUES FOR LAWYERS AND JUDGES
In most cases, a photographic
evidence speaks itself, but it needs a story. A picture generated by AI creates
problem in that scenario. For instance, an official at higher position may feel
oppressed due to the notion that he is committing a crime when, in fact he is
not committing such an act. It will be very challenging for the judiciary or
attorney that wants definitive proof of the scenario to conclude the truth
regarding the incident. Then, how do the courts possibly know that that picture
is indeed AI-made, not actual? Uncensored steps in the making of AI lack
transparency and prejudice in training could lead to inaccurate results;
numerous other risks make its evidence reliability into question. The lack of
established standards for proving AI-generated evidence complicates the
decision-making process. Governments of states are thinking about this problem
and working to enact legislation to address concerns related to generative AI.
Such leads that help identify the culprit beyond a reasonable doubt are termed
as evidences. However, it would be difficult in determining who is responsible
and to whom the criminals will hand over the responsibility if robots take over
human tasks with the help of their own "brain". Further, an
autonomous vehicle which could function even with no human control may be
difficult to consider electronic and AI-related evidences.
For example, it is not clear how the findings of a
drowsiness detector may be used in adversarial or inquisitorial legal systems
to determine who is responsible for an accident. Could this information be used
as evidence to determine the mastermind behind the accident? Since machines can
now think like humans, how will the concept of mens rea change? Could
human-machine interaction-based machine data be used as evidence? We have to
scrutinize the correctness and limits of the data set used by the AI system,
attribute blame to mishaps or disagreements, and understand the rationale behind
its decision-making.
THE ROLE
PLAYED BY JUDGES
Judges will be expected to gain adequate understanding about
how the AI works to take intelligent decisions as regards admitting evidence
into record that are created using such systems. Understanding must involve all
possible ways, which would make such admissible and others not to include how
it might misrepresent with its ability such as in the deepfake mode.
The purpose of the UNESCO article "The Admissibility
Challenge: AI-Generated Evidence in the Courtroom" was to familiarize
legal practitioners with the role of AI-generated evidence in judicial
processes and with the challenges likely to arise in determining whether such
evidence may be admitted or not. Such challenges include a question around the
reliability and authenticity of AI-generated evidence as well as the capability
of judicial actors in developing comprehension and evaluation competencies for
the technical aspects of the evidence. It further provides valuable information
regarding best practices to develop capacity among the judicial actors on
AI-generated evidence issues. As the AI technology is progressing, the
AI-generated evidence has become very common in the courts, but the legal
system still has a problem of how to handle this kind of evidence.
AI AS DATA
SECURITY TOOL IN LEGAL FIELD
AI can be used to scrutinize data in ways that were
previously impossible. For instance, AI can be used to analyze large amounts of
data and identify patterns or correlations that may be relevant to a factual
dispute. Lawyers have already seen this sort of tool used by e-discovery
vendors to assist in the initial review of large discovery productions in
complex cases. One area where proponents of AI may claim to be particularly
useful is in the analysis of audio and video recordings. Similarly, AI could be
used for the analysis of facial expressions and body language. However, it is
objected to misleading uses.
AI integration into the prosecution is
both an opportunity and a hindrance-in particular, the admission of evidence.
The picture isn't bright though; only a handful of court rulings have thus far
talked about the admissibility of AI-generated evidence, and most of these have
given no thought whatsoever to the issue. In contrast, admissibility of
evidence has always been determined by some important factors: fairness,
authenticity, relevance, and reliability. The introduction of AI-generated
evidence adds yet another layer of complexity: the AI operation would also need
to be explained to explain how it produced the evidence. The decision-maker
must know that basic operations of the AI system, how it functions, and how it
generates output so as to be in a position to accord the weight to the
evidence.
Under the Federal Rules of Evidence
(FRE), mostly, the trial judges are gatekeepers for determining the
admissibility of relevant evidence, including AI-generated content. The party
offering Admittedly AI-generated content must explain how it operates and how
the content will aid and not confuse the jury into a just verdict. Such
processes will involve revealing enough information concerning the training,
development, and operational features of the AI model so that both the opposers
and the judge could measure its reliability. Multiple factors play a pivotal
role in determining the admission and relevance of AI evidence. These factors
include the accuracy and reliability of an AI system, the interpretability of
the algorithms, and privacy invasions dictated by massive amounts of data.
Meanwhile, other considerations such as bias in the AI system, timing, and the
basic constraints of AI, can certainly influence the relevance and
admissibility of such evidence in court. The authenticity of AI-generated
evidence shall depend on showing, under FRE 901(a), that the evidence was what
it appeared to be. This is the point at which the jury is going to be largely
influenced by it. Discussions around witness testimony and corresponding systems
or procedures under FRE; clauses hinges on the provision of rules 901(b)(1) and
901(b)(9).[1] However, establishing the
authentication of evidence that uses AI is not an easy task. The areas of major
difficulty involve lack of transparency of the algorithms governing AI
technologies, the bias in training datasets, the provenance and quality of the
used data, how compliant the technologies in question are to be with the law,
and the absence of any legal understanding of artificial intelligence as technology.
These pose enormous barriers to authentication and thereby inherently question
the accuracy and reliability of the evidence.
RELIABILITY
OF AI GENERATED EVIDENCE
Programs leveraging
artificial intelligence, including but not limited to Gemini, CoPilot, and
ChatGPT, have radically changed our approach towards many facets of modern
life. It stands to reason that AI, as it is developed further, will likely be
called upon to play a pivotal role in more and more court cases. Just how
reliable will AI be in determining whether or not someone is telling the truth?
Just how accurate will AI be in reading body language and facial expressions? A
purely legal angle here, where it becomes incumbent upon lawyers to analyze the
data that AI looks at, works in an interesting way, because they might also get
themselves into analyzing the AI's own programming. Where in reality, an
expert-who is not denied to be costly-conducts the analysis instead of the
lawyer. In other words, while AI tool-just that-looks like a useful thing, it
tends to aggravate the already high cost of litigation for the average citizen.
The AI programs might themselves be error-prone. Like all software, AI is
fallible and error-prone. This means the evidence developed through AI may not
be reliable. The evidence may bump into some major hurdles against
admissibility.
CONSIDERING
AI AS EVIDENCE
Litigation in the present
context involves complicated business and intellectual property issues;
therefore, many worry about trade secrets, finance data, or other sensitive
information are continuing to mount. AI tools could simplify the examination of
this data in certain ways. Introducing this kind of analysis as evidence adds
another layer of complexity and cost. For instance, electronic "signatures
on other classes of documents may help prevent counterfeiting, but proving the
reliability of the electronic signatures themselves may need several levels of
testimony from experts. To prevent these challenges, courts will have to
develop new standards and protocols for the admissibility of AI-generated
evidence. This may involve setting guidelines on the use of AI in litigation
cases, including allowing litigants to review the source code of the AI
program. This may include establishing protocols for the disclosure or
certification of AI-generated evidence, or requiring independent third parties
to confirm the evidence before it is used in court. Judges may be unduly
influenced by the convincing nature of AI-generated content, especially if the
technology is not understood. For example, the realistic appearance of
deepfakes may deceive fact-finders and lead to miscarriages of justice. Thus,
the court system's ability to investigate and understand the underlying
technology is inextricably connected to the reliability of AI-generated
evidence. The court's rejection of AI-enhanced video evidence in State of
Washington v. Puloka[2] reveals the judiciary's reluctance toward
AI-generated information. The judgment raises issues regarding the inability of
existing legal frameworks to justly assess the validity and reliability of such
evidence.
Court as Admissible
The infusion of artificial
intelligence into the justice system presents a very revolutionary change
equipped with its share of innovations and dilemmas. Following a number of
scandals, in which some lawyers put before a court legal briefs sustained by
fictitious case citations fabricated on or through an artificial intelligence
mechanism, several judges have done their part of updating the standing orders
to address, specifically, how to deal with AI-generated content in filings.
However, as AI-produced evidence is increasingly becoming common and
indistinguishable from the non-AI variety, courts will also have to deal with
considerable and complex issues regarding the authenticity, reliability, and
admissibility of evidence generated by AI. Of serious concern to the courts is
the use of AI to create manipulated videos, images, or "deepfakes,"
that is, artificially produced images, video clips, and audio recordings that
are indeed fake but can pass for the real thing and that could taint an entire
trial. As an example, in the area of intellectual property, deepfakes could
illegally take advantage of a person's image and likeness and trademarks and
branding. If a deepfake should use copyrighted material, that might expose it
to copyright violations, and it might also stir some controversies about
authorship or inventorship issues. That very aspect will bear towards all these
controversies landing in AI-generated evidences in trials. Thus, within
intellectual property, a deepfake can be used to illegally present an
individual's image and likeness, trademarks, and labels. If deepfakes
incorporate copyrighted material, they may present claims of copyright
infringement and raise questions as to authorship or inventorship; thus, these
issues will certainly give rise to the introduction of AI-generated evidence at
trial.
The contribution of generative AI has undergone a radical
change regarding how intellectual property rights are traditionally viewed. The
development is going to have far-reaching implications for technology, the
economy, and society itself. Historically, intellectual property rights have
played their own unique roles in motivating creativity and safeguarding the
creations of its authors. The advancement in AI technology creates problems and
complexities for the traditional concept. The intellectual property remains
relevant today and is forming an increasing number of users. While some new
challenges are emerging, the approaches to these may involve adding IP
protection in layers, rather than expelling the attempts to modify IP
completely. This creates an interesting problem for the admissibility and
reliability of AI-created content in a legal context requiring new standards
and methods to establish such admissibility and reliability.
As the development of AI technology continues, so does the
progress of the legal framework to ensure that law and equity are wrapped up
with the other challenges raised by AI-generated evidence. In our time, AI
algorithms and machine learning models can autonomously bring forth various
creative content that contests our mutable understanding of creativity as it
calls for human ingenuity. The principle of exclusivity grants the right holder
an absolute power of exploitation that gives him an edge in fostering further
creativity. But the questions arise as to: who is the author? Who owns it? What
is the uniqueness of the work? Hence, finding a happy balance between
appreciating AI as an exercise in creativity and yet keeping a human-centered
balance is crucial.
CONTENT OF OWNERSHIP
The creative sectors might also be radically changed. The
onslaught of AI-generated material threatens to lower the value of art made by
human hands and disrupt the livelihoods of writers, musicians, artists, and
other creators, especially with a high volume and at a low price. This raises
questions about the disruption-and-replacement of human creativity from economy
by AI.
INFLUENCE
ON TRADE, PATENT, COPYRIGHT
According to WIPO's estimations, the progress in AI by this
industry has highly impacted intellectual property. The World Intellectual Property
Organization revealed that between 2013 and 2016, the average growth rate of AI
technology was 28%. Approximately 340,000 patent applications on AI-related
technologies and more than 1.6 million academic publications on the topic were
generated from 1956 to 2017. AI patent applications made to WIPO reached 55,660
in 2017, 300% up from 2011. Such developments have posed several challenges to
intellectual property law. In India and the USA, works created solely by AI
systems with no human input are not protected by copyright or patents.Thaler is
a Missouri-based powerhouse of innovation led by a CEO and president, Stephen
Thaler. It has spawned a movement with the foundational purpose of questioning
the status quo. Thaler is especially known for unleashing the awe-inspiring
DABUS technique, from which an artwork titled "A Recent Entrance to
Paradise" emerged using picture analysis of a gigantic collection.
The
DABUS case presents a considerable legal challenge to traditional patent
law by asking the question of what constitutes an "inventor." The
applications raised in this series of legal proceedings are of inventions
allegedly conceived by an artificial intelligence system named DABUS
("Device for the Autonomously Bootstrapping Unified Sentience").Patents
have been sought, mainly filed in various countries such as the United States,
Europe, and United Kingdom, by Dr. Stephen Thaler, developer of DABUS, who
chose to name DABUS the inventor. in the choices of different courts regarding
this matter warranted the courts to answer the question of whether DABUS could
be an inventor under the law in favor of Dr. Thaler. All the time, it
progressed under arguments that the existing law doesn't explicitly exclude AI
systems from inventorship. However, some courts ruled otherwise, with the UK
Supreme Court in Thaler v
Comptroller General of Patents Trade Marks And Designs[6] stating
that only natural persons could be named as inventors in patent applications.
The DABUS case raises deep questions about the nature of invention, the role of
human creativity, and the relationship between humans and AI in innovation. It
challenges the doctrine that only humans can be inventors and requires a
critical review of existing patent law to determine how such innovation driven
by AI will meet legal standards. It tests the future of intellectual property
law and guides the potential future impact that AI would have on human society.
In short, it will become a pivotal analysis to determine the content ownership
by testing far and wide elemental basics.
The
Getty Images v. Stability AI[7] case is a
high-profile legal fight that illustrates how complex copyright law has become
with the advent of AI. The case was brought by Getty Images, a prominent stock
photo agency, against Stability AI, which developed the well-known
text-to-image AI model, Stable Diffusion. The Getty Images' claims that
Stability AI trained Stable Diffusion on a massive dataset of copyrighted
images without obtaining proper licenses or permissions from Getty Images or
other copyright holders. Getty Images argues that this unauthorized use
of copyrighted material constitutes copyright infringement. They argue that
Stable Diffusion, by learning from and replicating patterns and styles found in
copyrighted images, effectively "copies" those images in a way that
infringes on the rights of photographers and artists. This case raises crucial
questions about the legality of training AI models on massive datasets of
copyrighted material, particularly when such training is conducted without
explicit consent or compensation to the copyright holders.
Though
not directly related to AI, the case of "Monkey Selfie" (Naruto v.
Slater) provides insightful understanding about the changing concept of
authorship. A photographer named David Slater had claimed the copyright of a
selfie of a crested macaque monkey, taken using Slater's camera. Arguably, the
case, again involving a monkey as the alleged author, raised again the
"issue of whether an animal could be an "author" for purposes of
copyright". Even though the court found in favor of Slater, the decision
showed that conventional copyright law was increasingly at a loss in
determining creators who were not actually human and called for further legal
and philosophical consideration of authorship into the age of digitalization.
While different in their particular contexts, these two cases touch upon the
very questions of authorship, ownership, and the application of copyright law
when technology is rapidly changing. The Getty Images v. Stability AI case
deals specifically with the important issue of AI-generated content and the
possibility of copyright infringement. The "Monkey Selfie" case
broadens the discussion to ask a question about the definition of
"authorship" outside the human world. These are great
precedents and also shape the new legal landscape pertaining to AI in
intellectual property.
INTERSECTING
AI EVIDENCE WITH IP LAWS
The
intersection of AI-generated content and AI-generated evidence poses a
particular myriad of legal and ethical challenges, especially with respect to
intellectual property rights (IPR). Though questions of authorship,
originality, and ownership relative to AI-generated content exist, its actual
evidence-like usage makes the situation more complicated. The capability of
artificial intelligence to generate deepfakes, manipulated videos, and
synthetic audio jeopardizes the authenticity and reliability of evidence. In
particular, advanced forgeries might make up false incrimination-related
evidence, incur the right of the testimony that would not otherwise exist, or
even create entire tales. It would damage the integrity of the proceedings, and
it brings up the, whether in cases of defamation, harassment, or political
manipulation.
In
addition, AI's employment for the analysis of evidence-such as forensic analysis
or document review-raises serious issues about the ownership and conclusion
control of generated data. With the application of AI systems on evidence
analysis, the intellectual property rights with the results and conclusion
could create confusion. For example, in the conduct of a legal discovery, who
then owns the copyright to the AI report that reviewed a large constellation of
documents Moreover, the possibility of AI algorithms being biased poses a
substantial threat. AI systems are largely based on data that is huge but has
intrinsic bias, which will likely be reflected in the output which will thus
conduct biased or inaccurate analysis on the evidences and in turn suffer from
the fairness and reliability of assessment during trial processes. For instance,
consider AI which has been trained on biased data that would, in turn,
attribute races to certain people to indicate their potential as suspects-this
could be seen as discriminatory.
The
situation has arisen due to the necessity for the multi-pronged approach being
increasingly implemented by the presence of AI-generated evidence within legal
and intellectual property frameworks. Strong mechanisms need to be introduced
for the authentication and admissibility of evidence generated by AI in courts.
This should necessarily include the specification of AI-oriented evidence
validation protocols, for example, those associated with forensic analysis
tools to detect deepfakes and fabricated content. Courts should insist on AI
expertise to assess the veracity of such evidence and also set up training
programs for judges and lawyers that will facilitate the better understanding
of the implications of AI technologies. Legislatures must also amend evidence
laws to be more particular in accommodating these new works created by AI so
that only content whose credibility has been verified is admitted before
courts.
As
regards intellectual property (IP) laws, only a radically reformulated
conception of authorship and ownership will temper the role of AI in content
production. The creation of categories in legal field such as "AI-assisted
works" or "machine-authored" can help bridge the conflict
between existing IP frameworks and the nature of autonomous AI functioning.
Moreover, in order to ensure fair ownership recognition for developers, users,
and data providers, policy guidelines should be made. These should aim at an
international framework for the harmonization of copyright and patent laws to
prevent inconsistencies and foster clarity, for example, as seen in DABUS.
Equally, greater accountability mechanisms must be put into force on AI
developers and users. The developers ought to ensure that their AI systems are
visible and audit-able so that people can trace the origins and purposes for
which they generate content. Users ought to be responsible when cobbling up bad
content that will mislead others. Independent bodies should look into AI
applications within the professions of law and creativity and institute redress
across the board for problems and encourage ethical AI practices. Reacting to the challenges regarding
AI-generated evidence and material will require a holistic approach involving
legislative intervention, technological innovation, and worldwide cooperation.
The
urge for legal reform following the rise of AI systems as creators and
contributors across different spheres of life has severely challenged the
traditional legal framework. It covered two major areas on these challenges:
admissibility and probative weight of AI-generated evidence in courts and
disruption of intellectual property laws by AI-generated content. Thus, a
combination of live cases and analyses demonstrates how AI deepfakes,
fabricating evidence, and art raise more nuanced questions with respect to the
principles of authenticity, authorship, ownership, and liability. With respect
to legal evidence, it is the very sophistication of AI technologies which in
turn complicates the courts’ task of
endorsing authenticity by distinguishing between what is genuine and what has
been fabricated. The already-existing laws primarily create hindrances for the
identification of AI-generated evidence considering originally when developed
for those by human beings. So do the laws of IP, derived from individual
humanistic concepts of authorship and originality, conflict with the dimension
of AI being more contributory with less human involvement than was sanctioned
before. Such a gap, anachronistic regression from new methodologies, is the
prime manifesto for much-needed reformation.
To
deal with these problems, a balanced approach using legal, technological, and
ethical solutions is imperative. The courts must adopt advanced mechanisms and
expertise with respect to the evaluation of AI-generated evidence;
restructuring of IP laws is necessary to make room for the much-invoked AI
systems which are contributing too much to the creative process.
Stakeholders-including policymakers, practitioners of law, and AI
developers-need to work together to ensure that the guidelines are unambiguous,
aligned, and equitable for all other parties concerned. In the end, however,
adjusting the law is a question for a new evolution to be conceived for
handling AI content, preserving the integrity of courts and intellectual
property control. Progressively approaching such reform can enable AI to
unleash its own revolution without any compromise on the standards of justice,
equity, and creativity.
·
Eftychia Bampasika, | ‘Artificial
Intelligence as Evidence in Criminal Trial’ (2021)
2844 CEUR-WS accessed on 23
January 2025
·
Maura
R. Grossman, ‘The GPTJudge: Justice
In A Generative Ai World’ (2023)
23(1) Duke Law & Technology Review
accessed on 23 January 2025
·
AI Rapid Response Team ‘Digital
Evidence and Deepfakes in the Age of AI’ (2024)
NCSC
accessed on 23 January 2025
·
Paul W. Grimm, ‘Artificial
Intelligence as Evidence’ (2021)
19(1) Northwestern Journal of Technology and Intellectual Property Northwestern
Journal of Technology and Intellectual Property
accessed on 23 January 2025
·
V. K. Ahuja, ‘Artificial
Intelligence And Copyright: Issues And Challenges’
(2020) SSRN accessed
on 23 January 2025
·
Abhishek Dalal, ‘Deepfakes
In Court: How Judges Can Proactively Manage Alleged Ai-Generated Material In
National Security Cases’ (2024)
SSRN < https://dx.doi.org/10.2139/ssrn.4943841> accessed on 23 January
2025
·
Sneha,
‘The Transformative Influence Of
Artificial Intelligence On Intellectual Property Rights’
(2023) 2(2) Symbiosis Law School, Nagpur Multidisciplinary Law
Review
accessed on 23 January 2025
[1]
Mavrova Heinrich, Denitsa and Pont, Erika, Who You Gonna
Call: The Role of Expert Witnesses in Authenticating AI-Generated Evidence
(2024).
[3] Bradley Budden, On the Intersection of Artificial
Intelligence and Copyright Law (2022), 47 CAN. L. LIBR. REV. 10
[4] Zack Naqvi, Artificial Intelligence, Copyright, and
Copyright Infringement (2020), 24 MIPR. L. REV. 15
Article Information
JUSTICE IN AI- GENERATED WORLD AND ITS LEGAL LANDSCAPES BY: SIDDHI JAISWAL
Authors: SIDDHI JAISWAL
- Journal IJLRA
- ISSN 2582-6433
- Published 2025/03/12
- Issue 7
About Journal
International Journal for Legal Research and Analysis
- Abbreviation IJLRA
- ISSN 2582-6433
- Access Open Access
- License CC 4.0
All research articles published in International Journal for Legal Research and Analysis are open access and available to read, download and share, subject to proper citation of the original work.
Disclaimer: The opinions expressed in this publication are those of the authors and do not necessarily reflect the views of International Journal for Legal Research and Analysis.