Published on 05th June 2025
Authored By: Kaustav Das Sharma
Sister Nivedita University
Introduction
AI and machine learning cause a revolution in multiple sectors all over the planet, including healthcare, finance, transport, and even the justice system. Such tech enables machinery to tackle tasks requiring human intellect, including decision-making, problem-solving, and recognizing patterns. AI and machine learning show great potential, but as they become part of our daily lives, they bring up big legal and ethical issues that current laws aren’t ready to handle.
The growth of self-driving cars and AI-powered medical diagnoses raises new liability issues when something goes wrong. Also, the widespread gathering and examination of personal info by AI systems brings up urgent worries about data protection and privacy. The chance for AI to keep algorithmic bias going makes the legal scene even trickier in areas like hiring and criminal sentencing. , the lack of thorough and clear regulatory rules creates doubt leaving key questions about who’s responsible and how open things should be up in the air.
This article will look into these key legal hurdles zeroing in on problems like liability, data protection biased algorithms, and gaps in regulations. It will also check out how current laws need to change to keep up with the quick progress in AI and ML tech. Plus, it’ll dive into why we need forward-thinking rules to make sure things stay fair, accountable, and safe.
Legal Background and Context
The growth of AI and ML technology has created many legal and ethical challenges around the world.
Global Perspective
A key legal challenge is figuring out who’s at fault. The EU’s AI Act tries to regulate AI systems based on their risk level. But there’s still confusion about who takes the blame when AI tech like self-driving cars causes harm[1]. The United States also grapples with liability issues with autonomous vehicles where figuring out who’s at fault in crashes involving AI gets tricky.
Another big worry is data protection. In the EU, the General Data Protection Regulation (GDPR) controls how companies gather and use personal data, including data that AI systems process. Yet, AI’s power to learn and handle huge amounts of personal data creates problems with openness and control[2]. In California, the California Consumer Privacy Act (CCPA) gives some protection to consumers but lacks broad federal laws to address all the ways AI’s data processing affects people[3].
Algorithmic bias presents a worldwide challenge. When AI systems learn from skewed data, they can reinforce or even boost existing social prejudices in areas such as recruitment and loan approval. The EU AI Act tries to tackle this problem by demanding openness and equity for AI systems that pose high risks[4].
Indian Perspective
The Personal Data Protection Bill, 2019 (PDPB) tries to tackle data protection issues. The Bill wants to control data collection and processing, but lawmakers haven’t passed it yet. This leaves holes in the legal shield for personal data[5]. The Consumer Protection Act, 2019 sets up rules to deal with harm from AI systems. However, it doesn’t zero in on AI-specific problems like biased algorithms or machines making choices on their own[6].
When it comes to biased algorithms, India’s National Human Rights Commission has voiced worries. They’re concerned that AI might treat people especially in areas like credit scores and hiring[7].
Key Legal Challenges
AI and Machine Learning are causing a revolution in many fields, from healthcare to finance, education, and law enforcement. These new technologies offer big benefits, but they also create many legal challenges that we need to solve to make sure AI is used and . These challenges include liability issues, data protection concerns, algorithmic bias, and gaps in regulations. They’re becoming important for lawmakers, government officials, and companies to think about. This article looks at these challenges from both a global and Indian point of view.
- Who’s Responsible When AI and ML Systems Mess Up?
The issue of liability continues to be one of the most urgent legal hurdles for AI and ML. As AI systems gain more independence, the standard liability rules that apply to human actions and faulty products no longer suffice. When an AI system inflicts damage—be it through a data leak, bodily harm (like in self-driving cars), or money losses—who bears responsibility?
Global Perspective:
In the EU, the AI Act[8] suggests regulating AI based on how risky it might be for people’s health, safety, and basic rights. But the question of who’s responsible when things go wrong is still up in the air. This becomes a real head-scratcher when AI systems do things on their own. We’re left wondering: Should the blame fall on the people who made the AI, the folks using it, or the AI itself? Take self-driving cars, for example. We still don’t have clear answers about who’s at fault when accidents happen.
Indian Perspective:
India’s legal system doesn’t yet tackle AI-specific liability. Still, the Consumer Protection Act, 2019 and Product Liability Laws might offer ways to deal with damage from AI systems. The Personal Data Protection Bill, 2019 (PDPB) zeroes in on data-related harm from AI systems but doesn’t address bigger questions about who’s responsible[9]. As AI keeps changing, India’s laws will need clearer rules on who’s liable and accountable for harm caused by AI.
- Data Protection and Privacy Concerns
Big datasets are a prerequisite to the proper functioning of an AI system, and these commonly include sensitive personal data. Obtaining, using, and processing such personal information brings forth major considerations relating to privacy and the protection of those rights. Hence, data protection laws are important in governing the processing of personal data with regard to AI.
Global Perspective:
In the EU, the General Data Protection Regulation (GDPR) acts as a strong legal structure to protect personal data. It has rules that tackle issues AI and ML create Article 22. This article gives people the right to avoid automated decision-making without human input when these choices affect their legal rights[10] Even with these safeguards, AI’s power to process huge amounts of personal data makes it hard to follow GDPR rules. This is true because many AI systems work in ways that are hard to understand.
Indian Perspective:
India’s Personal Data Protection Bill 2019 (PDPB) seeks to control how people gather and use personal data. It pays attention to data subject rights and the right to be forgotten. But the bill is still being looked at and hasn’t become law yet. As AI tech gets better, India’s laws need to deal with worries about AI data use. They must make sure data protection laws stay strong and cover everything[11]. India doesn’t have one big AI law, which makes things hard for AI companies that handle private info.
- Algorithmic Bias and Discrimination
One of the biggest legal hurdles we face is algorithmic bias. AI systems pick up on patterns from past data, which can reinforce and even boost existing societal prejudices. We see unfair results in areas like hiring, criminal justice, credit scoring, and healthcare showing how AI can add to systemic unfairness.
Global Perspective:
In the EU, the AI Act tackles algorithmic bias by making high-risk AI systems go through impact assessments to check their fairness and openness. The GDPR also has rules about transparency making sure people know the reasoning and importance of AI-based choices[12]. But making sure these rules are followed is still a big task. In the United States, people have filed lawsuits and taken regulatory actions against biased AI systems especially in predictive policing where algorithms have been shown to target certain communities more than others.
Indian Perspective:
In India, people worry about unfair computer programs in areas like credit scores and job searches. The National Human Rights Commission (NHRC) points out that AI systems can make unfair choices when they learn from biased information hurting groups that already face problems[13]. But India doesn’t have specific laws or rules to deal with unfair computer decisions in AI systems. As more people use AI, the country needs to create laws that stop unfair results.
- Missing Rules and the Need for Complete Laws
AI is changing faster than current rules can handle. Gaps in regulations are one of the biggest problems in AI law, as existing laws often can’t account for how quickly technology improves.
Global Perspective:
The EU has made a big move with the AI Act, which tries to regulate AI systems high-risk ones. But these rules might struggle to keep up with new AI tech. The lack of a worldwide system for AI rules makes this problem worse. Different places are making their own AI rules, which leads to a split approach to managing AI[14].
Indian Perspective:
India acknowledges the responsibility of AI and has implemented regulations as the guiding principle of the National Strategy for Artificial Intelligence, issued by Ministry of Electronics and Information Technology (MeitY), among several important points. India has made some moves toward AI regulation through the National Strategy for Artificial Intelligence, which the Ministry of Electronics and Information Technology (MeitY) published. Yet, this strategy focuses more on developing AI than regulating it. With the PDPB and other related laws under review, India still has a big regulatory gap in AI governance[15]. As AI tech keeps changing, India will need to create complete AI laws to make sure people use these technologies .
Discussion and Critical Analysis
The elaboration of AI and ML provides enormous advantages but also raises gigantic legal and ethical concerns. One important such concern relates to liability: classical liability frameworks are grossly incapable of examining harm done by autonomous AI systems, such as self-driving cars or automated financial trading systems. As such, the huge legal gap in the question surrounding who is liable for an AI-induced harm-that is, the developer, the user, or the AI-will increasingly be of great concern. While some of these issues are being attempted to be addressed by the EU AI Act and India’s Consumer Protection Act, 2019, vagueness remains. Future reform of existing laws would, however, need to articulate accountability mechanisms as the technology advances, with more sophisticated and autonomous AI systems[16]
Data protection, another major concern, also somewhat overlaps with the earlier issue. In fact, AI relies heavily on huge sets of data, some of which contain highly sensitive personal information. While some strong protection comes under the aegis of the GDPR in the European Union, it is still a matter of contention upon its applicability upon AI, especially upon automated decision-making. The PDB, 2019 in India talks about data privacy but does not provide guidance or solutions for AI-based processing of data; this highlights the chasm of regulation for data laws specifically aimed at AI[17]. As AI technologies keep evolving, it is high time for India to empower its legal system to attain effective data protection mechanisms to ensure individual rights.
The last but by no means the least is algorithmic bias that has often afflicted AI systems from the onset of their training. Discrimination in hiring, criminal justice, and credit scoring can aggravate social inequities. The European Union AI Act and reports by the National Human Rights Commission raise alarms about these biases; however, combating them is no longer just a question of finding technical solutions. There must be broad regulatory scrutiny and algorithmic transparency to prevent discrimination[18]
Conclusion
AI and ML technologies present vast opportunities, but they also introduce profound legal challenges that require urgent attention. AI systems’ ethical deployment revolves around issues of liability, data protection, algorithmic bias, and regulatory gaps. While some such as the EU’s AI Act and India’s Personal Data Protection Bill refer to bring proposed solutions to these challenges, they often fail to provide effective answers [19] [20]The question of liability in AI remains contentious, with the existing structures for statutory liability and tortuous liability being inadequate for the intricacies posed by AI technology, especially when dealing with autonomous systems. The issue of algorithmic bias cries for an even deeper scrutiny of the transparency, fairness, and accountability in AI systems, as the very reason behind creating algorithmically biased decision-making should be to avoid reinforcing existing inequalities in societies. Data protection legislation in Europe, especially the GDPR, is the strongest possible legal structure for guiding AI systems, but these are slowly eroding as the field moves fast with its developments, especially automated decision-making[21]
There needs to be stronger and more specific regulations made to meet these issues head-on while walking step by step with AI development. The world’s governments must then work to develop comprehensive frameworks to deal with such new issues, with India being no exception. Only when such a strong regulation exists can we safely deploy AI and ML without undue pressure on rights and social justice.[22]
References
[1] European Commission, Proposal for a Regulation on Artificial Intelligence, COM(2021) 206 final (legislative proposal)
[2] European Parliament, General Data Protection Regulation (GDPR), 2016/679 (regulation)
[3] California State Legislature, California Consumer Privacy Act (CCPA), Cal. Civ. Code §§ 1798.100–1798.199 (state legislation)
[4] European Commission, Artificial Intelligence Act (EU legislative proposal)
[5] Personal Data Protection Bill, 2019, The Parliament of India (draft bill)
[6] Consumer Protection Act, 2019, The Parliament of India (act)
[7] National Human Rights Commission, “Human Rights and Artificial Intelligence” (report)
[8] European Commission, Proposal for a Regulation on Artificial Intelligence, COM(2021) 206 final.
[9] Personal Data Protection Bill, 2019, The Parliament of India.
[10] European Parliament, General Data Protection Regulation (GDPR), 2016/679.
[11] Personal Data Protection Bill, 2019, The Parliament of India.
[12] European Parliament, General Data Protection Regulation (GDPR), 2016/679.
[13] National Human Rights Commission, “Human Rights and Artificial Intelligence”, 2019.
[14] OECD, Artificial Intelligence Principles, 2019.
[15] Ministry of Electronics and Information Technology (MeitY), National Strategy for Artificial Intelligence, 2018.
[16] European Commission, Proposal for a Regulation on Artificial Intelligence, COM(2021) 206 final.
[17] Personal Data Protection Bill, 2019, The Parliament of India.
[18] National Human Rights Commission, “Human Rights and Artificial Intelligence”, 2019.
[19]European Commission, Proposal for a Regulation on Artificial Intelligence, COM(2021) 206 final.
[20] Personal Data Protection Bill, 2019, The Parliament of India.
[21] European Parliament, General Data Protection Regulation (GDPR), 2016/679.
[22] AI for Social Good, ‘The Challenges of Dealing with Legal Issues Related to Artificial Intelligence’ (5 December 2023) https://aiforsocialgood.ca/blog/the-challenges-of-dealing-with-legal-issues-related-to-artificial-intelligence accessed 16 April 2025.
