Laws For Artificial Intelligence and accountability for its malfunctioning.

Author: Shantanu Gupta, NMIMS Navi Mumbai

Artificial Intelligence (AI) is a computer science discipline that enables machines to perform tasks that demand human intelligence. The activities include experience learning, human language comprehension, pattern recognition, and decision-making. Technologies in AI employ advanced technologies like Machine Learning (ML), which enables machines to learn and perform better as a result of experience based on an analysis of data. Natural Language Processing (NLP) enables AI to understand and respond to human language, and computer vision enables machines to recognize and understand images and video.

AI is used widely in all areas; i.e., in healthcare to diagnose disease, in finance to detect fraud, and in transport to develop autonomous vehicles. Siri and Alexa are also examples of common applications of AI.

There are two general categories of AI: Narrow AI, employed to perform a specific task, and General AI, employed to think and behave like a human being (although this is in the pipeline currently). Although AI is useful in many respects, e.g., efficiency and accuracy, there is the matter of privacy, job loss, and ethics of justice and bias.

Artificial Intelligence (AI) has proved to be of great worth to most industries and has simplified the life of humans. The best thing about AI is that it can work more efficiently and faster than human beings. For instance, in the manufacturing and healthcare sectors, AI can do repetitive work and eliminate errors. It is not just time-saving but also enhances the quality of work.

The second benefit of AI is that it never feels fatigue and is able to work 7 days a week and 24 hours a day. Human beings get fatigued easily, but AI does not require any rest. This is an extremely useful feature in customer service where chatbots are able to serve customers around the clock. AI is even capable of processing huge amounts of data and recognizing patterns that human beings tend to miss. This allows companies to make intelligent choices and optimize their plans.

AI is also utilized in hazardous conditions such as mining or space travel where human life is at stake. AI robots can conduct such operations without endangering human life. AI assists physicians in identifying patients with ailments and determining the patient's well-being, which results in suitable treatment and rapid recovery.

Apart from this, AI supports our daily life by giving personalized recommendations on websites such as YouTube, Netflix, and shopping websites. It knows what a user prefers and recommends accordingly. AI is also a major element in cybersecurity because it detects and blocks cyber-attacks.

In education, AI provides individualized learning experience to the students in the form of virtual teachers and smart learning platforms. It enables the students to learn at their own pace and improve the understanding of the concepts. 

In general, AI makes everything simpler and more efficient, more accurate, safer, and reduces the cost and enhances services in most sectors.

Artificial Intelligence (AI) is transforming the world, but it also has a list of negative consequences on society and individuals.

The biggest issue is unemployment. Robots and AI computers are able to do functions faster and more efficiently than human beings. In the manufacturing sector, for example, machines replace employees, and in customer service, chatbots respond to customer queries. This leads to unemployment and financial problems for many people.

The second problem is that it is extremely costly to construct. Constructing and maintaining AI systems is extremely costly. It relies on extremely advanced technology and extremely skilled individuals. Small businesses or developing nations might not be in a position to buy AI systems.

AI also lacks the creativity and emotional sense of human beings. Although AI can do things according to data, it cannot think creatively or feel like human beings. For instance, an AI program can assist medical professionals in diagnosing diseases, but it cannot emotionally console patients.

Data privacy and security are two of the most important issues. AI systems need a vast amount of data to function. However, this data can be hacked or used improperly, leading to privacy violations and security risks. For example, personal information stored in AI systems can be hacked into.

AI systems themselves may also be discriminatory and biased. If an AI system's training data is biased or skewed, the system will be biased. Facial recognition systems, for example, have been found to be biased against specific races or genders.

Another limitation is excessive dependence on technology. The more work is handled by AI systems, the more dependent humans become on them. This diminishes human imagination and critical thinking power. In case of a failure in an AI system, humans might not be able to accomplish things on their own. AI also poses ethical issues. For one, AI may be applied to weapons or surveillance networks, which are injurious to human beings. There is also controversy regarding the extent of control one should relinquish to AI systems in decision-making processes that have significant implications.

Lastly, nobody is held responsible. If an AI system made a mistake or caused harm, it is difficult to hold anyone responsible – the company, the creator, or the AI. This raises legal and ethical issues.

Briefly, while there are many advantageous uses of AI, there also exist critical challenges like job elimination, privacy, and ethics. In order for these negative consequences to be averted, these should have supportive regulations and measures.

India lacks a definite direct law that particularly talks about Artificial Intelligence (AI) exclusively. Yet, India is slowly applying laws and regulations such that AI systems can be used in the right manner and be brought to justice for what they perform.

1. National AI Strategy (2018)

The government of India, via NITI Aayog, introduced the #AIForAll campaign. The policy has been guided through the implementation of AI in the most significant sectors of healthcare, agriculture, education, and transportation. It has also remained dedicated to developing ethical AI, data privacy protection, and security with the aim of preventing misuse.

2. Sector-Specific Guidelines

In Banking and Finance: Reserve Bank of India (RBI) has come out with guidelines for AI-based online lending platforms. According to the guidelines, customer data is safeguarded and AI-driven decisions are transparent and fair.

In Healthcare: The National Digital Health Mission is concerned to highlight the necessity of robust regulations in implementing AI to detect and treat diseases. There should be an assurance of patient safety along with appropriate medical advice. 3. Digital India Act (Draft Bill - 2023)

The Digital India Act, being a new bill to supplant the current IT Act 2000, will incorporate the latest technologies of Artificial Intelligence. The Act will provide standards for data, security protection, and for the employment of AI in a judicious manner. AI firms will be held more accountable by law for the actions of their AI systems.

4. Judicial Actions by Courts In 2024, there was a legal victory for popular Bollywood actor Anil Kapoor against individuals who had copied his voice and image through AI without his consent. The case established the precedent that AI systems could not use an individual's identity for exploitation and that criminals would be prosecuted according to the law.

5. Data Privacy Law (Digital Personal Data Protection Act - 2023) This policy protects the private data of users, which is essential for AI systems that collect and process data. It keeps companies that use AI in check by imposing tough regulations that preserve users' privacy. 

There is no unique law in India to govern AI, but the government is developing strategies and policies in various industries. Because of the intersection of digital laws, data protection culture, and judicial rulings, India is trying to make AI accountable, equitable, and secure.

The current legal structure in India, such as the National AI Strategy of 2018, The Digital Personal Data Protection Act of 2023, and the pending Digital India Bill of 2023, safeguard against AI misuse to some extent. Nonetheless, these legislations cumulatively lack their ability to mitigate AI accountability. The primary justification being the legislation which concentrates on data privacy and cyber security, passively faces challenges like discrimination, unethical conduct, and damage from independent systems. One major conundrum which stems from the autonomous betrayal is the lack of ambiguity about who should be accountable for the mistakes and damages needlessly perpetrated whilst using the AI system. Is it the designer of the system, the user, or the AI system itself? AI is not fully regarded as a legal entity under Indian law, thus placing all the blame on the AI itself is not an option. Consider the case of an Autonomous driven car that meets with an accident. Is it the car company that gets the flag or the software engineer who designed the bloody thing or the helpless driver? In such cases, the manufacturer should be held liable for designing AI that is safe to be around, and the user should be called out for operating it proficiently.

But there also needs to be a system in place to track and audit the AI systems on a constant basis to avoid mistakes and prejudice. The European Union and the United States are implementing laws that address these issues more proactively.

India needs legislation and a structure of accountability to monitor AI misuse. 

To begin with, India is poised to enact the AI Liability Law which will determine the extent of damage an AI system is allowed to injure. For robo cars, the law has to provide clarity on whether the automaker of the vehicle or the operator is liable in the event that the car hits a pedestrian. In the same spirit, AI systems can be scrapped by an AI Ethics and Safety Law that sees to it that no safety and equity issues are embedded into the design of the system. This will be done by making it mandatory for developers to work on the periodic audit of the AI systems for bias and embedded error cleansing. 

The next step would be an AI Accountability and Transparency Law that will bring to the open the black box AI companies hide their systems behind. A bank that makes use of AI to grant or refuse loans should be able to tell customers the logic the AI uses in making decisions. Also imperative is an AI Data Protection Law which will defend unconsented use of personal data by AI systems.

To ensure there are new and better ways of improving supervision, auditing and enforcement of these offending systems, a very crucial step to consider is that the government should establish an AI Oversight Body, together with everything else that needs to be taken care of. There should not be any anthropocentric approach where the systems are allowed to cross a boundary.

Through its efficiency, accuracy, and automation features, Artificial Intelligence (AI) is changing the world for the better. On the contrary, these transformations also come with their share of challenges, including displacement of employment, ethical problems, invasion of privacy, and gaps in accountability. Although the National AI Strategy (2018), the Digital Personal Data Protection Act (2023), and the proposed Digital India Bill (2023) create some level of regulation within India, It still lacks frameworks to address AI specific problems like bias, liability, and autonomous decision making. 

Robust AI-specific legal frameworks are necessary for ensuring responsible AI implementation in India. This should include guidance AI Liability Law which clarifies responsibility in AI related harming incidents, , AI Ethics and Safety Law which calls for audits beyond the expected bias or safety checks, AI Accountability and Transparency Law which obliges explanation pertaining to AI decisions made, and AI Data Protection Law which prevents unauthorized usage of personal data. Furthermore, designated AI Supervisory Authority would streamline monitoring and enforcement for the AI compliance orders. 

The India case illustrates that it is possible to create a stiff structured legal framework that causes AI innovation to behave ethically and not have adverse effects on people and society.