Open Access Research Article
ARTIFICIAL INTELLIGENCE AND ITS CORRESPONDING EFFECT IN LEGAL SYSTEMS
Published Paper
Article Details
ARTIFICIAL INTELLIGENCE AND ITS
CORRESPONDING EFFECT IN LEGAL SYSTEMS
AUTHORED BY - SUDIP MUKHERJEE
sudip35ee@gmail.com, 9064423085
LL.B. from R.P.S. Law College, Patna, affiliated to Patliputra University
Legal Intern at District &
Sessions Judge Court, Bankura
Abstract: -
This research work will try to
enlighten impact of Artificial Intelligence (AI) in the legal systems, its
various prospects on the everyday legal practice and its over the years performances used by legal practitioners,
lawyers, advocates or pleaders or mukhtars or barristers. This paper also finds
out adverse effects of AI all over the world especially in India. Unlimited
uses by billions of people over the years tells a wholesome story of AI and its
impact in the legal systems. Previous backdrops were removed and some
challenges are being faced by the developers and new developed technologies are
used for betterment of legal practice. Some issues are solved and some are
still grave for the years to come. The legal practice is being largely affected
by the emerging platforms of AI month or month or day by day gradually and at a
very high speed. AI is no-doubt becoming to be one of the most controlling
organ in the path of legal practice and it will be definitely grasping the
baton from the human in reality and legal system is no different from this influence
of AI.
Background :- [1]
It is of some significance to mention The
TV mini-series Class of 09 which tells an interesting story
about the influence of AI on the criminal justice system in the USA. And in the
last episode, the AI accuses an advocate for her anti-AI views in her
unpublished book. During the trial, The advocate argues in trial for deciding
her guilt or innocence before the judges because the judges panel decided AI to
be perfect in terms of calculations. Though our present situation still lacks
dependency on AI but it will no-doubt be the largest independent platform for
the judicial systems in future years.
Recently, every
practice, profession or trade whatever the industry be is somewhat fluctuated
by one name that is for sure AI. Our
Indian Government though have limited the scopes of AI in legal practices in
its ambit still today due to regulations and predominant contracts, which are
being the controlling measures for implementing AI in larger perspective. Now
AI have limited uses in automated contract review, legal research,
transcription services, etc.
Usages and or Benefits of AI in the legal profession
·
In The Matters of Law Firms & Lawyers [1] [2]
AI offers
manifold benefits before the lawyers either senior or emerging new faces to
help them in drafting and also increasing the efficiency, reducing the costs of
different legal works and giving a platform for more compact and robust work
environment for legal practice. AI can handle many kinds of routine tasks and
mechanical works such as document and contract review, legal research and data
analysis. Law firms are growing on this aspect of AI day by day at the instance
of increasing productivity and profitability. However, AI is still not yet
capable of handling more complex tasks such as deal structuring, negotiation,
advocacy and representation in a courtroom. The regular use of AI may definitely
decrease costs of law firms in respect of hiring legal inters or lawyers. Rich
or big firms can easily implement costly AI systems while the small firms
cannot do so.
·
Indian Judiciary Systems
[2]
In the year
2021, the Supreme Court used an AI-based tool which could process data/ information
and make that information available to judges but it does not participate in
the decision-making process. Another tool that is used by the Supreme Court of
India i.e. SUVAS (Supreme Court Vidhik Anuvaad Software) which basically translate
English legal papers/works into different vernacular languages and vice versa.
In the
case of Jaswinder Singh v. State of Punjab, the Punjab &
Haryana High Court rejected a bail petition due to allegations from the
prosecution that the petitioner was involved in a brutal fatal assault. The
presiding judge requested input from ChatGPT to gain a wider perspective on the
granting of bail when cruelty is involved. However, it is important to note
that this reference to ChatGPT does not express an opinion on the case's
merits, and the trial court will not consider these comments. The reference was
solely intended to provide a broader understanding of bail jurisprudence when
cruelty is a factor.
Usage of AI in the judiciary: Background and Uses in
Different Countries [1]
o
USA
In USA, AI-powered
tools such as COMPAS (Correctional Offender Management Profiling for Alternative
Solutions) is prominently used to help judges in analyzing factors such as
criminal history, socio-economic background and mental health of the accused.
The US Sentencing Commission also utilizes AI to create and enforce sentencing
guidelines for fair and just punishment.
Chatbots
are used vastly by the US court system for answering faqs (frequently asked
questions) about court procedures, schedules and other related subjects effective
for information to the citizens. USA Government try to reduce the work-pressure
of the staffs related to courts and apt response to public grievance and easy
access to basic information regarding to courts.
o
CHINA
China's
Smart Court System is totally based on AI technologies which can compare to
retrospective studies or background check and also can give suggestion relating
to which law or precedent will be applicable to the concerning case or not.
This system can also recommend sentences based on same types of cases, and thus
giving the judges to deliver justice rapidly.
Chinese
courts give preference to AI systems for legal research. The 'China Judgements
Online' platform, powered by AI, offers the citizens a quick response towards
their queries.
o
UK
The UK Ministry of Justice introduced
the Digital Case System in 2020 for the crown courts which actually gives
updates in real-time and remote court participation and also allows for the
digital submission of evidence which in turn reduce usages of papers. The Bar
Council's Ethics Committee provides their guidelines for criminal law barristers for accessing
the online portal.
Legal framework available Globally and in India [1]
Though AI
can benefit the society in terms of health, education, transportation and
entertainment etc, yet on the other hand there are negative sides as well such
as security threats, violation of privacy and so many.
To cope up
with the growing day new versions of AI in market, a new
voluntary framework is released for the usage and development of AI
securely. The World Ethical Data Foundation (WEDF) has 25,000 members,
including staff at tech giants such as Meta, Google and Samsung. The framework
contains 84 questions for developers to consider at the start of an AI project.
There are
white papers, guidelines and policy in jurisdictions such as UK, USA and EU
which target algorithmic impact assessment and elimination of algorithmic
biases. The European Parliament recently adopted amendments to its proposed
Artificial Intelligence Act. The new amendment of European Parliament is trying
to ban the use of AI technology in biometric surveillance except for law
enforcement, subject to judicial authorization, and for generative AI systems
like ChatGPT to disclose AI-generated content.
Uses in India [1]
Unfortunately, no specific guidelines or laws are there in
India to regulate the use of AI. Ministry of Electronics and information
Technology (MEITY) [8], is the
executive agency for AI-related strategies and had constituted committees to
bring in a policy framework for AI.
The Indian Government had built a new platform, Niti Ayog
which has instructed for a set of seven responsible AI principles, which
include safety & dependability, equality, inclusivity and
non-discrimination, privacy and security, transparency, accountability and the
protection and reinforcement of positive human values. According to Indian
Constitution, The Supreme Court and High courts in India have a constitutional
mandate to enforce fundamental rights one of them is the right to privacy. In
India, the Information Technology Act was amended up to date and it is the only
act which is related to data privacy and punishments for the breach of it.
Besides, the Digital Personal Data Protection Bill has been introduced by MEITY
[8], though its enforcement is still
a far cry. This bill if become effective one, citizens will have the right to
ask for the reasons why the data is being collected from them by either private
or government authorities and many more.
Risks and challenges of uses in legal practice [1]
Data Privacy
AI systems usually depends on sets of data to adapt
themselves and act accordingly. So one can imagine the risks as those data may
have some information relatively sensitive in nature, such as personal or data
regarding one’s financial information. Now there is a chance that those data
could be gone in the wrong hands.
Bias
Potential bias is one of the very common risk and
challenge of in AI systems. The AI systems may not collect actual history or
may mal-function in its programming and thus resulting in demerit of the
systems it is attached with.
Licensing and
Accountability
AI systems, without proper license can be dreadful
afterwards as there will not be subjected to any legal regulations, legal ethical
standards and professional codes of conduct. Now one simple question – “Who
will be responsible if AI mislead with some wrong or improper calculation or
information? Or its developer will be under the hands of law?”
AI imposes a real threat to the actual judiciary in
decision making in future years and relying on AI can confuse on the supremacy
of Judiciary in One Country’s all over scenario.
As per one recent New York Reporters, a New York lawyer
used ChatGPT for legal research and included six case citations in a brief
filed with the court and surprisingly opposing counsel could not find any of
the cases, and afterwards the law firm associated with the concerned Lawyer had
to pay the fine of $5,000 in total as per fine imposed by the Judge. As a
result, the lawyers have to be immensely cautious relying on the ChatGPT or
other AI based platforms.
Competition
As per framework established by the Competition Act, 2000,
AI systems had undergone relevant huge competitive bases over the years all
over the world. India is not also far behind.
Establishing error-less technologies or human like
capability of cognition or brains developing is a real challenge for the
developers and scientists in the matters of accountability in case of AI
systems.
Needless to say that AI could not be a substitute for the
lawyers by any means, may it be a mere complimentary but not real human-being.
AI cannot take very realistic decisions upon studying papers or data very
complex in nature.
At the end of the day, the advocates or lawyers will be
liable for their cases/works and they have to protect their client's interests
and documents at any cost. A lawyer's expertise and experience cannot be anyhow
compared with that of AI’s.
ChatGPT [3]
According to OpenAI announcement 2022, ChatGPT is a
sibling model to InstructGPT, which is trained to follow an
instruction in a prompt and provide a detailed response.
Firstly, OpenAI tried ChatGPT to get users’ feedback and
learn about its strengths and weaknesses. During the research preview, usage of
ChatGPT is free and then launch official website namely chat.openai.com.
Methods [3]
ChatGPT basically uses Reinforcement Learning from Human
Feedback (RLHF), using the same methods as InstructGPT, but with slight differences in the
data collection setup. The initial model used supervised fine-tuning: human AI
trainers were provided with conversations where they were playing for —the user
and an AI assistant, both the sides. Composing responses, overall mixing new
dialogue dataset with the InstructGPT dataset, and in the end transforming into
a dialogue format – this is the main method used for ChatGPT.
For reinforcement learning, gathering comparison data is
a must, which was made of two or more model responses ranked by quality. In the
way of collecting comparison data,
conversations were used for AI trainers have with the chatbot. Random selection
of a model-written message, sampled
several alternative completions, and afterwards them ranked by the AI trainers
and then these reward models are used for fine-tuning the model using a
method called as Proximal Policy
Optimization. Performing
several iterations of this process is all about ChatGPT.
ChatGPT is fine-tuned from a model in the GPT-3.5 series,
which finished training in early 2022. ChatGPT
and GPT-3.5 were trained on an Azure AI supercomputing infrastructure.
§
Limitations of
ChatGPT [3]
§ ChatGPT
sometimes gives incorrect or nonsensical answers.
§ ChatGPT is
sensitive to tweaks to the input phrasing or attempting the same prompt
multiple times.
§ The model
is often excessively verbose and overuses certain phrases, such as restating
that it’s a language model trained by OpenAI.
§ Ideally,
the model would ask clarifying questions when the user provided an ambiguous
query. Instead, current models usually guess what the user intended.
§ Though
OpenAI tried using the Moderation API to warn or block certain types of
unsafe content, but it is still giving some false negatives and positives for
now.
Iterative deployment [3]
Latest
research release of ChatGPT is a very unique step in OpenAI’s iterative
deployment of
increasingly safe and useful AI systems. Many lessons from deployment of
earlier models like GPT-3 and Codex informed the safety mitigations in place
for this release, including substantial reductions in harmful and untruthful
outputs achieved by the use of reinforcement learning from human
feedback (RLHF).
ChatGPT in India [3]
OpenAI
recently launched its AI-powered chatbot, ChatGPT, and those are readily available
as app version for download on Android devices in India. Earlier, the app version
was only accessible to iPhone users. Just like the web version available online,
the app allows users to ask the chatbot questions on any topic and receive
conversational answers.
Model Case Study:-
Simply
imagine a Courtroom or Chamber of an Advocate applying all sorts of latest
techniques of AI. Let’s take A, a person came to B’s Chamber. B is a Lawyer of
a District Court in India and he is using AI system for data collection. B is
starting to give information about his legal damages related to property which
actually is just a narration of how his brother, C is not giving his permission
to partition the ancestral property with all possible means. A is speaking in a
vernacular language which is programmed in AI i.e. say ChatGPT generator D,
used by B. But the problem arises when B tries to read the data stored in
ChatGPT. Actually ChatGPT though took all the responses by A but could not
decode some words properly and at last the case study by B was not so
successful as he has to carry out routine procedure to use his own expertise
and experience to take all the information given by A. Later A was out of town
for some days and his Law Clerk, E uses the ChatGPT generator D for gathering
some information and unintentionally sent some confidential data to another
WhatsApp Number instead of his own WhatsApp Number. After few months B’s
personal account was hacked and his all property was transferred in the name of
some unknown person, F.
Here
are some points to be noted:- [4]
i)
The Lawyer, B must
take more care towards his AI interface, GhatGPT generator, D and privacy and
other related confidential details about his any client/clients.
ii)
The Lawyer, B must
not rely blindly over the AI systems.
iii)
The AI authority
should make more robust and safe interface so that nobody can easily break the
security interface of AI.
iv)
Cyber Authority
should come out with some new techniques to track this kind of data theft.
v)
No Person should be
allowed in an Advocate’s or Lawyer’s chamber in his/her absence especially
regarding any Confidential information of Clients.
vi)
Trainings and campaigns
should be provided for the importance of AI Systems, ChatGPT techniques
available.
Remedies:-
There
are some provisions of the Laws available in any Country i.e. Cyber [5] Laws must be enforced in every
possible way for our help in our Systems.
AI
uses and its certain security measures cannot be compromised at any cost and
for that some proposed methods can be given as under:-[4]
1. Cyber Cells
must be associated with the cases of AI related problems and queries in every
Police Station at least two or three as per the area span of the district
concerned.
2. Cyber Cell
Authority is to be appointed on utmost expertise, skill and talent combined
with experience and they all must follow the rules of law and procedures as per
specified under the Sections of Cyber Law.[6]
3. Necessary Trainings
should be provided to the persons associated with the Cyber Security [7] in any
Department of Administration, Legal or others.
4. Bank
Authorities or Internet Services Cafe centers or any legal authorities should take
utmost responsibility in the cases of AI related Crimes.
5. AI should
be used with proper care and nobody should rely fully on AI.
6. ChatGPT
method should be provided with all sorts of security measures needed in its
path.
7. Online
awareness about ChatGPT, any AI interfaces and its all side-effects should be
given more importance.
8. Ethical
Hacking courses are increasing inside the Country but this effort should not be
wasted at any cost and should be accordingly the sections of the Cyber Laws.
9. Courts
should deal with the growing cases of Cyber Crimes more and more cautiously and
thus it can be in some way helpful in cases of AI related crimes.
10. The
Advocates should be provide adequate information and necessary trainings and
documents, guidance in using AI interfaces, ChatGPT etc.
11. The Laws
should be amended time-to-time as the AI is and will be changing everyday or
may be every seconds afterwards.
12. The
Government should pass bills or notifications necessary and websites should be
provided with up-to-date notifications of each and every new provisions or
alterations in Cyber Laws and as of AI.
13. Last but
not the least, self-sufficiency is very important. If not fully possible to
cope up with the modern trends then at least be aware of spam emails or ads, or
jump to any sort of conclusions without prior thinking or not to indulge in
sharing of any confidential information at any cost without proper guidance
from the concerned authority.
Conclusion:-
If
there is internet on the run, there will be misuse of it by someone for sure.
But one should not blindly or recklessly share any data to anyone unknown or
known without valid reason or without any reason to be provided so. Cyber Laws must
have to be enforced in every aspects to strengthen the measures of Cyber Cells
and other Authorities regarding Cyber Activities. AI is very new horizon of
Internet Activities and ChatGPT is one of the most spoken interfaces of OpenAI
so far. So, handle with care is the most needed thing to do in case of
using AI in our lives otherwise it can be
detrimental to all of us.
Bibliography:-
[1] Artificial
Intelligence in context of legal profession and Indian judicial system by Aditi
Prabhu.
[2] Artificial Intelligence and its
impact on legal-technology by Sterling Miller.
[4]
Cyber Security In
The Light of Present Day Un-authorized Uses- International Conference Proceedings
(15th Virtual) on Right to Privacy In Digital Era – Challenges And
Solutions Published and Organized by Kerala Law Academy Law College in
Association with CALSAR, Trivandrum, 17th December, 2022 Author: Sudip
Mukherjee,
[5] Cyber
Crimes and Laws – Virtual Book by Sushma Arora, Raman Arora.
[6] Handbook of Research on Cyber Crime and Information Privacy
(2 Volumes) Maria Manuela Cruz-Cunha (Polytechnic
Institute of Cávado and Ave, Portugal) and Nuno Ricardo Mateus-Coelho (Polytechnic
Institute of Management and Technology, Portugal)
[7] Cyber Crime Investigations
Bridging the Gaps Between Security Professionals, Law Enforcement, and
Prosecutors 1st Edition - March 21, 2007 Authors: Anthony Reyes,
Richard Brittson, Kevin O'Shea, James Steele
2021.pdf
[9]
https://infosecawareness.in/cyber-laws-of-india
Article Information
ARTIFICIAL INTELLIGENCE AND ITS CORRESPONDING EFFECT IN LEGAL SYSTEMS
Authors: SUDIP MUKHERJEE
- Journal IJLRA
- ISSN 2582-6433
- Published 2023/09/27
- 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.