Fundamental Rights in the Looming AI-Powered Future: Challenges and Opportunities

I. Introduction

The capabilities of artificial intelligence (AI) systems have rapidly expanded in recent years as a result of better data accessibility, higher hardware performance, and improved AI model performance. Specifically, the release of so-called large language models (LLMs), such as Google Bard and Chat GPT, has spurred a wide-ranging societal discussion about the ethical and human rights implications of AI. One AI's creation and use are already causing significant social upheaval. In addition to law enforcement and the criminal justice system, this is having a significant impact on the mobility, health, education, and energy supply sectors. However, there are instances where the technologies employed are the focus of contentious discussion. Examples include automated analysis of personal or health data, the creation of facial recognition software, and the usage of lie detectors. Businesses are crucial to this issue since they have a role in both the creation and application of these artificial intelligence (AI) technologies. The analysis of customer data, CV evaluation, production use of AI, plant safety use of AI systems, and algorithm-based consumer willingness to pay assessment are all pertinent fields. As a result, companies using AI must increasingly confront the technology's human rights implications." This is partially because of the increasing adoption of global business and human rights norms, which are fueled, among other things, by the UN Guiding Principles on Business and Human Rights and the UN Global Compact. It is also a result of stricter laws governing AI.(Chatterjee et al., 2022)

II. The definition and application of ai

The literature does not yet have a common definition of artificial intelligence. The phrase describes the ability of computer programs or algorithms to mimic intelligent human behavior, such as operating a car, identifying pictures, or producing written language. "Outputs such as content, predictions, recommendations, or decisions influencing the environments they interact with" are what the EU AI Act defines as AI's outputs in Article 3(1). A variety of methods and ideas have served as the foundation for the development of AI. Deep learning and reinforcement learning are two types of machine learning (ML) that are currently the subject of special attention. Based on experience and input from the surroundings, both strategies allow AI systems to learn on their own and make decisions on their own:

• The application of deep learning in object and picture recognition is growing. Algorithms in medicine that examine CT scans or X-ray pictures for questionable patterns, such cancer metastases, are one particular application.
• AI systems that create their own autonomous approaches to problem-solving are referred to as reinforcement learning. The method, which is often based on a system of rewards or penalties, is being utilized more and more in robots and autonomous driving.

These illustrations show how AI can be applied to a wide range of spheres of life. Robots, language systems, and analytical tools are a few examples. Additionally, this technology has uses in the police, military, and intelligence sectors. As a result, AI is a multipurpose technology with frequently unexpected applications. The employment of text analysis software or the assessment of psych demographic information by law enforcement are two examples. (Chatterjee et al., 2022)

III. Artificial intelligence and human rights

AI affects people's daily lives in our lives and exhibits an overall propensity to turn choices, recommendations, etc. into computation. It is thought to possess "intelligence" in the sense of being able to predict the future and provide a concrete solution to a simple, challenging problem in a more straightforward manner. AI technology is demonstrated by a variety of gadgets, including laptops, tablets, smartphones, drones, robots, and even self-driving cars. People utilize these gadgets for a variety of tasks, including everyday domestic chores. Moreover. Police investigations and even military operations benefit from these gadgets. Anything that can be coded may be done by algorithms if the necessary data can be accessed.

A design structure can then be articulated, and this structural frame is useful for carrying out any necessary activity. It is a huge advancement in this regard. If it is possible to guarantee the availability of data covering all potential human activity and a number of other operations, the algorithmic ability is boosted out of proportion. According to the definition given in the previous section, this data is referred to as big data. Accurate predictions about the future can be made with the aid of AI technology in the form of machine learning. This is accomplished practically by ML's analysis of the pattern. It is undeniable that algorithms operate more effectively than humans. But biases still exist in the algorithms. They are known to promote human biases. Human biases are present in systems that were created by humans. Furthermore, this data are used to build algorithms, which are products of human design. Algorithms may, of course, contain biases unless the proper precautions and controls are implemented. At this point, it is necessary to clarify what "bias" means. From an AI standpoint, "bias" can be described from both a statistical and a sociological perspective. "An inclination or prejudice for or against a person or group, especially, in a way that is considered to be unfair" is how bias is interpreted at the societal level.

The bias may be examined differently in the context of statistical complexity. Bias is defined as the discrepancy between the actual (true) value and the expected value of a system or value in this context. There is another explanation for it. It is the distinction between what the system predicts and what actually occurs. In the majority of cases, it is seen that each AI system's statistical bias produces something, and that "something" introduces societal bias. Human rights issues, such as the need to protect human life, are always brought on by this prejudice in AI's capabilities.

A person who favours one concept over another and typically does not give it the same chance is said to be biased. Bias is affected by the quantity of components. Popularity, benefits, favouritism, etc. are some of them.

(A) Why Human Rights Always Matters

AI has often disproportionately affected the most vulnerable and poor, and has "created new forms of oppression." The idea of human rights tackles power imbalances and gives those who support them the vocabulary and processes to challenge the practices of more powerful entities, such as governments and corporations. Human rights are universal and legally binding, and they are established in a body of international law. While governments have additional responsibilities to uphold and protect human rights, both businesses and governments must respect these rights. The application of human rights law to evolving situations, including advancements in technology, is clarified by a vast network of regional, national, and international institutions and organizations that offer well-developed frameworks for remedies.

Furthermore, the moral legitimacy of human rights has substantial normative power in situations when domestic legislation is deficient. Identifying and condemning human rights abusers is frequently a useful tactic, as violating human rights has political and reputational repercussions on a worldwide scale. Regulations pertaining to human rights have the potential to reduce some of the most severe social problems caused by artificial intelligence and prevent similar effects in the future.(By et al., n.d.)

(B) General Risks Of AI

The specific fields of application (like banking or medical) are directly related to the human rights dangers that come with the usage or development of AI. Nevertheless, regardless of their particular function, AI systems also have some general traits that pose hazards to human rights. Combining clever algorithms with vast amounts of data is a fundamental aspect of AI systems. Since the use of biometric and medical data as well as the exploitation of information from social networks raises concerns about data protection, this presents a number of human rights challenges. AI systems that can identify persons by their speech or perform in-depth personality analyses are two examples of this. AI is also capable of drawing conclusions that are not fully foreseeable throughout the development stage. This is because the model may change in response to each new piece of information that an AI system receives. Because of the opacity or lack of explainability that results, humans find it challenging to comprehend the decisions that AI takes. However, unexpected errors that have serious human rights implications may result from the AI model's features or the data set's lack of representativeness. Because certain groups frequently lack access to adequate quality data, or may not have it at all, inadequate data input can result in people being treated unfairly. Vulnerable populations (such those with a migratory background) have historically been disproportionately impacted by so-called algorithmic biases. In particular, this may entail that some demographic groups must pay more for airline tickets or require later medical treatment appointments than the general public. The trend toward standardization is another facet of the application of AI. Typically, AI cannot take the role of a single person, such as a physician making a diagnosis. As opposed to that, it alters entire specialties, like cancer diagnoses, or entire ecosystems, like hospital operations. In most cases, AI-related errors are not isolated incidents or anomalies. Therefore, the essential terms opacity, complexity, dependence on data, and autonomous behavior are used in the explanatory brief on the EU AI Act to summarize the fundamental issues with AI. Humans can have an impact on all four factors, such as the choice of which particular data to use, the level of autonomy granted to persons involved in decision-making, and the degree to which decisions are standardized or customized.(Artificial Intelligence and Human Rights Recommendations for Companies, n.d.)

(C) How AI Impacts Human Rights

One of the main topics in the current ethics discussion is the well-documented role that AI plays in enabling prejudice. In order to acknowledge these problems, Access Now collaborated with AI firms and human rights groups to publish "The Toronto Declaration" in March 2018. However, AI implicates other human rights besides the right to be free from discrimination. AI has an impact on almost all internationally recognized human rights because they are interconnected and dependent on one another. We look at a number of the ways AI affects human rights below. The rights included are mostly those found in the three documents known as the "International Bill of Human Rights," which serve as the foundation for international human rights legislation.” Among these are the International Covenant on Civil and Political Rights (ICCPR), the Universal Declaration of Human Rights (UDHR), and the International Covenant on Economic, Social, and Cultural Rights (ICESCR). The European Union Charter of Fundamental Rights' right to data protection is an addition to these in this study. We address the concerns brought up by hypothetical future advancements in AI and also how current AI applications either infringe or have the potential to violate each relevant fundamental human right. It's crucial to remember that the following discussion of human rights concerns extends beyond artificial intelligence, or AI. AI's capacity to recognize, categorize, and discriminate increases the likelihood of human rights violations on a larger scale, despite the fact that there are already many in the area of digital rights. Similar to the human rights abuses linked to other digital applications that exploit data, the issues surrounding the usage of AI sometimes disproportionately impact vulnerable people. Women and children, members of specific racial, ethnic, or religious groups, the impoverished, people with disabilities, and LGBTQ people can all fall under this category. The data and results perpetuate the long-standing marginalization of these populations, reinforcing historical trends.(– European Union Agency for Fundamental Rights, 2020)

1. Impact On Right to Life, Liberty, And Security, Equality Before the Courts, A Fair Trial.

The criminal court system's increasing use of AI runs the risk of violating people's right to personal liberty. One example is the employment of deterioration risk-scoring software in the criminal justice system of the United States, which is used to guide decisions about detention at almost every level, from criminal punishment to the assignment of bail. Based to the program, more Black defendants have been unjustly categorized a greater threat, faced more stringent bail requirements, held pending trial, and sentenced prior to longer jail terms. Also, due to risk-scoring algorithms are not legally required also may include arbitrary inputs, detention decisions based on them may not be lawful. Criminal risk evaluation software is promoted as an instrument that judges can use exclusively to assist in sentencing determinations. However, determining a defendant's level of future responsibility by categorizing them as having an elevated or decreased likelihood of repeat offenders could compromise the presumption of truth required for an impartial prosecution. The utilization of historical data by predictive policing software can also include preexisting police bias, which could lead to an inaccurate assumption of guilt. Despite the fact that many courts rely heavily on the results due to the software's supposed objectivity, reports show that judges are usually unaware of how these risk-scoring algorithms work. Whether court decisions based on this software can truly be considered fair is called into question by this. By using these instruments, governments essentially give commercial contractors the power to make decisions. The engineers employed by these suppliers, who aren't elected officials, code policy decisions that are frequently hidden from the public and government agencies using data analytics and design decisions. When someone is denied parole or sentenced to a specific amount of time for reasons they will never understand and that the government body in charge of making that decision is unable to explain, their rights may be infringed and trials may be unfair. The inability artificial intelligence to manage nuances will likely lead to further problems in the future. There are situations in which breaching the law is acceptable; laws are not always absolute. For example, it is usually acceptable to run a red light to prevent a rear-end collision with a car that is tailgating. Red-light cameras can't tell the difference between a ticket and a traffic violation, but a human law enforcement official can determine whether to issue one. In a future where artificial intelligence powers intelligent cities and "robocops," this lack of delicacy could result in a sharp rise in the number of people who are unfairly detained, ticketed, or fined without any way to appeal. These conditions may eventually lead to a society where people prefer to carefully abide by laws and regulations, even in the face of extraordinary events, and lose the capacity to make critical decisions. It is predictable that as more and more data about our lives becomes available, AI-based systems that assist with legal and law enforcement judgments will integrate social networking posts and activity. Words or behaviors that suggest a violent inclination or the risk of committing particular types of violations could be identified using machine learning. The rights to equality before the law and a fair trial would also be violated by such use.(HUMAN RIGHTS IN THE AGE OF ARTIFICIAL INTELLIGENCE, n.d.)

2. Impact On Democracy, And the Rule of Law

Technologies related to artificial intelligence (AI) provide a number of chances to enhance both the way government and human lives operate. Artificial intelligence's strength, scale, and quickness can increase efficacy and efficiency in a variety of fields, including public administration, healthcare, education, and transportation. They can replace human labour in jobs that are hard, hazardous, unpleasant, and challenging. However, AI technology has the potential to undermine democratic institutions, the rule of law, and human rights. We should be better able to comprehend these combined advantages and risks since AI is "socio-technical"—that is, a broad range of sophisticated machines that operate within interpersonal contexts and are designed to accomplish people-defined goals. As a result, one may argue that AI systems reflect the values and choices of their developers and users. Artificial intelligence (AI) can be used to forecast behavior among people, spot medical symptoms, and evaluate threats to the welfare or interests of individuals. The welfare, opportunities, and rights of the people to whom these tasks are applied could all be impacted. For such systems to be developed and implemented, accountability is therefore essential. Despite the fact that AI can perform complex or laborious tasks in place of humans, decisions made during the development and implementation of AI systems may replicate harmful bias and other human judgmental fallibilities that negatively affect affected people and the population as a whole in manners that are more difficult to recognize than when carried out by living things.

Therefore, in order to be responsible for AI, we must carefully analyze the potential risks and advantages for people and communities in addition to assessing the technical components of a given system or technology. Inhumane bias, which can manifest overtly when AI systems predict prejudiced results or treat a certain identity or demographic category differentially than others without cause, is one of the possible negative effects. The opaque character of some AI systems complicates the assessment of their potential for damage. Artificial intelligence (AI) innovations may be challenging to understand or explain because of their technological intricacy and intellectual property rights, in addition to being created with specialized knowledge.

The provisions of the European Convention on Human Rights (ECHR) and the European Social Charter (ESC), which include particular guarantees pertaining to liberty and justice, privacy, freedom of expression, equality and non-discrimination, and social and economic rights, can be used to examine the particular human rights implications for AI systems. AI's effects on democracy and the rule of law also fall outside the purview of the ECHR and ESC, but they are nonetheless equally significant in their own right. We can decide where existing liberties and rights are adequate to protect us, where extra elaboration of existing liberties and rights is necessary, and where new liberties and rights must be tailored to the particular dangers and opportunities presented by AI and machine learning by carefully examining the dangers and possibilities posed by artificially intelligent systems.(The Public Policy Programme at The Alan Turing Institute Was, n.d.)

3. Impact On Privacy

Data exploitation: Consumer goods, such as linked cars, smart home gadgets, and phone apps, are frequently designed to exploit data. Consumers often face an informational imbalance in terms of the kinds and amounts of information produced, processed, or exchanged by their devices, networks, and portals. It is more important than ever to educate the public about this kind of data exploitation as we integrate more intelligent and connected gadgets into our homes, workplaces, public areas, and even our bodies.(Chaturvedi & Law, n.d.)

Identification and tracking: Artificial intelligence (AI) applications may identify and track individuals in public spaces, at work, and on a variety of devices. For example, AI can be used to de-anonymize personal data, which undermines the distinction between personal and non-personal data that is the foundation of current data protection legislation, even though personal data is commonly (pseudo-)anonymized within datasets. Facial recognition is another technique for identifying and tracking people, which has the potential to alter the norms around anonymity in public places. Around 69% of demonstrators who hide their faces with scarves and caps can even be recognized by machine learning systems. When it comes to law enforcement, face recognition technology can help officers identify people without the need for probable cause, reasonable suspicion, or any other legal requirement that might otherwise be necessary for them to obtain identification through more conventional methods.

Inference and prediction of information: Highly sensitive information can be anticipated or inferred from non-sensitive data types using machine learning techniques. Keyboard typing patterns can be used to anticipate people's emotional states, such as confidence, anxiety, melancholy, and fatigue. Profiling presents serious privacy issues and can lead to discrimination where private information that is confidential, such political beliefs, gender identification, cultural background, or well-being, may be predicted , from unrelated data (such as activity logs, phone metrics, location data, or social media likes).

4. Impact On Political Participation

"Advances in artificial intelligence and cyber capabilities will open opportunities for malicious actors to undermine democracies more covertly and effectively than what we have seen so far," according to a Brookings Institution paper. Researchers have highlighted Russian disinformation tactics using automated bots on social media as attempts to sway the 2016 U.S. presidential election. It may become increasingly challenging to identify internet bots that are being used as weapons to disseminate false information in political debate since AI is being developed to imitate human conduct in online chats.One of the many beneficial uses of bots is to assist search engines in finding material. However, those created with malevolent intent, such disseminating false information, have been discovered on social media sites like Twitter, undermining the likelihood of an informed populace, which is necessary for significant democratic elections. The ICCPR, for example, mentions the rights related to political activity in the entitlement to freedom of choice and the freedom to equitable representation in governmental and societal affairs. According to this human rights perspective, when a bad actor co-opts an automated system, it generates a human rights responsibility that needs to be addressed. But one of the most controversial topics in platform technology right now is picking the best strategy. Platforms are more likely to remove bots because they violate their terms of service than to do so in order to protect users' right to political participation. Further research would be necessary to ascertain how human rights notions might impact legal or contractual conflicts in this scenario.(Latonero, n.d.)

5. Impact on Freedom of Expression

In a world where social media sites employ algorithms to determine whose viewpoints we hear, the right to freedom of expression is especially crucial. An emotion al-contagion study was carried out in 2014 by Cornell and Facebook academics to look into how emotions spread over social networks. Through the use of sentiment analysis, the researchers were able to change the experiences of almost 700,000 Facebook users by identifying whether friends had posted or left negative remarks.

In order to investigate if algorithmically distorting the feed to show positive items might maintain users on the site more time, those negative postings were then deleted from users' newsfeeds. According to this study, platforms may choose how to display the world depending on user expressions, strengthening some realities while weakening others. The freedom of expression is one of the fundamental human rights protected by the Universal Declaration of Human Rights and Article 19 of the ICCPR. The role of social media platforms in content moderation is a hotly debated topic as these platforms have emerged as the primary venue for public discourse. Lawmakers and the public are urging businesses to address the issue of hate speech, fake news, and media manipulation that are spreading on social media sites like Facebook and Twitter. Concerns regarding how private corporations can effectively define speech boundaries are raised in response to these calls to action. According to David Kaye, The special rapporteur of the United Nations on the right to freedom of thought and speech, content filtering systems may inadvertently stifle minority perspectives and other unpopular but essential forms of free expression. While advocating for greater regulation of tech platforms, the NYU Centre for Business and Human Rights notes that government interference can sometimes backfire. Making tough choices about standards and their subsequent implementation is necessary due to the costly resources and constant labor of content management. According to Kaye, placing rights of human at the forefront of this discussion provides states, businesses, and other important parties with useful guidelines for regulating material in this hazy context. These guidelines would also be applicable when implementing automated systems. A human rights viewpoint tells us that not all rights for humans are absolute, even though this paper cannot go into great detail on this controversial topic. Proportionality, or weighing the legal and social effects in relation to various rights, is a factor in trade off judgments. Language for analyzing the harmony of other liberties with the freedom of expression, including those related to political involvement, information, assembly, association, privacy, and security, is provided by a rights-based framework.(Privacy and Freedom of Expression in the Age of Artificial Intelligence Privacy and Freedom of Expression In the Age of Artificial Intelligence Privacy and Freedom of Expression in the Age of Artificial Intelligence, 2018)

IV. Framework for protecting human rights in ai driven world

The human rights framework has shown itself to be adaptable and timeless, and it may be used to address modern issues. Digital rights, or those that grant us access to, use of, creation of, and publication of digital media, as well as the ability to access and use computers and communications networks and determine what information qualifies as personal data and, consequently, the right to privacy, are an example of how human rights are applied to contemporary issues. In addition to protecting privacy, digital rights also safeguard freedom of expression, association, and assembly. A new international framework to protect digital rights was not established after the swift digitization and information and communication technology advancements had unexpected effects on people's rights and liberties. Human rights and digital rights attorneys, on the other hand, only cited existing conventions and treaties to argue that human rights must be upheld both online and offline. But this is a continuous struggle. There is still much to be done to clarify the extent of digital rights and guarantee that freedoms and rights are safeguarded by international law "whether exercised in person, through technologies of today, or through technologies that will be invented in the future." These safeguards are also crucial for directing the advancement of artificial intelligence. Many ethical constraints are addressed by creating AI in accordance with human rights standards and holding it responsible for defending people's fundamental liberties and rights. We can build on the worldwide agreement that currently exists on human rights. With United Nations Member States31 bound by conventions and treaties crafted over decades, the human rights framework is one of the most pervasive international legal systems in existence, although not being entirely universal. Human rights law is well-defined, has precedents and a legal foundation, and is enforceable globally. These three qualities would greatly aid the development of AI, particularly in light of the fact that it has harmed people more and more in recent years, causing irreparable harm. Not only do signatories have international responsibilities, but many have also incorporated these treaties into their national constitutions. Following ratification, states pledge to protect, respect, and fulfill the rights and freedoms of all people within their borders, including protection from harm that could be inflicted by outside parties, as well as to incorporate human rights into their domestic domestic legislation. States must be able to verify that the power accumulation in private domestic businesses is legal. But this has turned out to be challenging. Big Tech corporations' economic, social, and political power increases as they influence domestic AI-related activities more and more each year. Both domestically and globally, states ought to hold them responsible. Furthermore, states may be able to govern in ways that violate citizens' fundamental rights thanks to the technologies developed by these commercial entities. The international community must hold states accountable in a same manner. Human rights can offer a foundation for upholding accountability and responsibility in AI technologies in a way that the ethical discourse is now failing to accomplish sufficiently well because they are enshrined in international law and institutions.(Saslow & Lorenz, 2019)

V. Conclusion

The escalating influence of artificial intelligence underscores the urgent need for a fortified international framework that integrates human rights into its ethical and operational core. Although the human rights regime—enshrined in international treaties and domestic legal systems—offers a strong foundation, the rapid evolution of AI technologies has outpaced regulatory mechanisms. This disparity allows states and corporations, particularly dominant Big Tech entities, to exploit AI advancements with minimal accountability, often infringing upon fundamental rights. The principles embedded in the human rights framework, characterized by their universality, enforceability, and legal precedents, present an unparalleled opportunity to bridge this governance gap. By mandating adherence to human rights standards in AI design and deployment, states and the international community can ensure these technologies serve as tools for empowerment rather than oppression. Governments must reinforce their obligations under international law by incorporating digital rights into domestic legislation, addressing data privacy, freedom of expression, and protection from technological abuses. Concurrently, mechanisms to hold private corporations accountable are imperative to curb their disproportionate influence on societal structures. Moreover, international cooperation is essential to prevent states from leveraging AI for authoritarian practices, such as surveillance and censorship, which erode civic freedoms. Human rights serve as a compass to navigate the ethical complexities of AI, transcending mere theoretical discourse to create actionable, binding standards. The integration of human rights into AI governance is not merely aspirational but imperative to preserving human dignity in an AI-driven world. By fostering an equilibrium between innovation and justice, the global community can safeguard liberties for present and future generations.

*****

VI. References

• European Union Agency for Fundamental Rights, F. (2020). Getting the future right – Artificial intelligence and fundamental rights. https://doi.org/10.2811/58563
• Artificial Intelligence and Human Rights Recommendations for companies. (n.d.). By, E., Tripathi, S., From, S. B., & Alliance, A. (. (n.d.). VIS-À-VIS MODERN INDIA. www.judicateme.com
• Chatterjee, S., N.S, S., & Hussain, Z. (2022). Evolution of artificial intelligence and its impact on human rights: from sociolegal perspective. International Journal of Law and Management, 64(2), 184–205. https://doi.org/10.1108/IJLMA-06-2021-0156
• Chaturvedi, A., & Law, I. (n.d.). Indian Journal of Integrated Research in Law DEFINING LEGAL RESPONSIBILITY IN THE AGE OF AI: ADDRESSING GAPS IN DATA PRIVACY REGULATION.
• HUMAN RIGHTS IN THE AGE OF ARTIFICIAL INTELLIGENCE. (n.d.). Latonero, M. (n.d.). Governing Artificial Intelligence: UPHOLDING HUMAN RIGHTS & DIGNITY.
• Privacy and Freedom of Expression in the Age of Artificial Intelligence Privacy and Freedom of Expression In the Age of Artificial Intelligence Privacy and Freedom of Expression in the Age of Artificial Intelligence. (2018).
• Saslow, K., & Lorenz, P. (2019). Artificial Intelligence Needs Human Rights.
• The Public Policy Programme at The Alan Turing Institute was. (n.d.). https://www.turing.ac.uk/research/research-programmes/public-policy.

*****