Abstract

The blog critically examines the impact of Artificial Intelligence on social justice by studying hindrance in systematic procedures called as “algorithmic bias.” It critically scrutinizes the current digital law frameworks including the European and Indian Laws and its implementation around the world with the help of real-life examples, advocating a need for adopting an equally protected approach for the future. It further explores the use of Artificial Intelligence in various fields that are sensitive to biases and prejudice for instance it being used for hiring by various Multi-National Companies such as Amazon and giving assistance to courts in the process of decision making and maintaining its administrative efficiency with the help of tools like The Correctional Offender Management Profiling for Alternative Sanctions (COMPAS), Harm Assessment Risk Tool (HART), Supreme Court Vidhik Anuvaad Software (SUVAAS), Supreme Court Portal for Assistance in Court’s Efficiency (SUPACE) etc, in a way that it should be in a structure to make people feel more just, equitable and inclusive.

Detailed Analysis

Artificial intelligence is one of the biggest creations of the 21st century. The mere existence of AI comes with its pros and cons, and one of them is the effect of algorithmic bias on the social order of the society, which tends to create discrimination on the basis of sex, caste, class, colour, religion, etc. The implementation of the European Union’s Artificial Intelligence Act^[i] is a huge global experiment; it is the world’s first major attempt to write a rule-book for the technology that is redesigning the entire society at an unbelievable pace. The EU act divides the AI systems into four major categories according to the level of risk they possess which includes unacceptable risk, high risk, limited risk and minimal or no risk^[ii]. Penalties for non-compliance are the highest and can go as high as EUR 35,000,000 or 7% of the worldwide annual turnover of the company, whichever is higher.^[iii]

The need for such a stringent law arises because of the algorithmic bias created by artificial intelligence systems. It occurs when systematic errors in a computer system create an unfair outcome, like privileging one outcome over another on the basis of social and societal differences^[iv]. It can give overwhelming results in sensitive areas such as criminal justice, healthcare, hiring, etc.

Its presence is clearly visible in big multinational companies like Amazon, which uses an AI tool for hiring that discriminates between male and female candidates; developers trained the hiring algorithm using resumes of former hires who were largely male.^[v] The Correctional Offender Management Profiling for Alternative Sanctions (COMPAS) is an AI system used by US courts to assess the defendants’ re-offending risk^[vi]; but it met with criticism when discrimination was suspected between Black and white defendants^[vii]. Similarly, problems were seen in the AI system HART, or Harm Assessment Risk Tool, adopted by UK courts, where biased discriminatory patterns were observed that racially profile people for recidivism risk.^[viii] One of the main reasons behind this phenomenon is the personal biases of developers, which affect the algorithm and cause it to behave in a particular manner. Another can be the wide-ranging datasets on which these systems are trained, which often duplicate biases present in the original datasets.

India’s integration with AI technology is very fast-paced, with the government investing in programs such as IndiaAI Mission, BharatGen AI, Yuva AI for All, etc.^[ix] But with the massive growth of artificial intelligence, the country faces a threat of algorithmic bias, which can affect the very sensitive social order of the nation. There has been a rise in discrimination based on AI technology, which can lead to violent results. For instance, police used AI based facial recognition Technology system to identify and arrest dozens of protesters after communal clashes in Delhi’s Jahangirpuri area in 2022. Most of the people charged were Muslims^[x], which led to criticism of India’s AI-based technology by tech developers, activists, and human rights foundations. The move was seen as a structured strategy to target the poor and people coming from a particular community by integrating artificial intelligence in a crucial domain.

In the last few years there has been a swift increase in the use of artificial intelligence in the justice system of the country. This move is seen as a way to reduce the burden on the courts with the help of steps like the launch of the E-court project in the country and the integration of first-generation pilot systems such as the Supreme Court Vidhik Anuvaad Software (SUVAAS) and the Supreme Court Portal for Assistance in Court’s Efficiency (SUPACE)^[xi]. The main aim of the first-generation pilot project is to improve the administrative efficiency of the courts and help in domains such as judicial process, legal research, case management, and law enforcement. The systems are tried under the monitoring of the AI Committee, whose main job is to ensure the smooth functioning of AI systems and check for the systematic biases present in the datasets. With the ongoing implementation of phase 3 of the E-Courts Project, the second-generation pilot system has taken the integration of AI systems to another level with the addition of predictive analysis in case outcome.^[xii] This move can significantly reduce the workload of judges, but if not effected with proper measures, it can lead to devastating results. The risk of influencing judicial discretion and undermining judicial independence comes hand- in-hand with the overreliance on predictive tools. India with vast backlog of pending cases can take its first major but calculated steps, it has to learn from the model adopted and measures taken by countries like Estonia, which has been fairly successful in absorbing the artificial intelligence model in their justice system by taking innovative steps such as designing a ‘robot judge’ to adjudicate small claims disputes of less than €7,000 (about $8,000). The pilot was initiated to resolve contract disputes and is aimed at eventually expanding to other claims.^[xiii]

India lacks a proper policy for tech companies and developers to compulsorily check for biases present and occurring in the data. Moreover, there should be centralized domain-specific laws to deal with the misuse of artificial intelligence in the country. This new age discrimination may not be heard, but it can be felt as we are moving towards an era of artificial intelligence. In a developing country like India, where the majority of the people come from a rural and marginalized background, individuals are already facing systematic tangible discrimination from the beginning. The inequality or bias created by artificial intelligence in various crucial facets of work will surmount their struggle in numerous forms and make them feel alienated in their day-to-day life in these modern times. Moreover, it will create a fear in the minds of people about new technological changes and hinder the process of digital growth which is equal and transparent in nature.

Conclusion

The dynamic nature of artificial intelligence presents it as an inter-generational life-changing technology, but concern arises when this tool tends to create a hindrance in equality between people through the means of algorithmic bias. Steps like the inclusion of artificial intelligence in the social lives of people in sensitive areas should not be taken in a hurried manner; rather, they should be in the form of a slow, long-drawn process with the inclusion of all stakeholders at the time of policy formation Irrespective of there social and economic background. As India and the world progress on the path of artificial intelligence, it is important that the inherent fear of discrimination in the hearts and minds of people should be mitigated rather than increasing.

^[i]https://artificialintelligenceact.eu/ai-act-explorer/

^[ii]https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai

^[iii]https://www.europarl.europa.eu/topics/en/article/20230601STO93804/eu-ai-act-first-regulation-on-artificial-intelligence

^[iv] https://www.datacamp.com/blog/what-is-algorithmic-bias

^[v]https://www.reuters.com/article/world/insight-amazon-scraps-secret-ai-recruiting-tool-that-showed-bias-against-women-idUSKCN1MK0AG/

^[vi] https://researchoutreach.org/articles/justice-served-discrimination-in-algorithmic-risk-assessment/

^[vii] https://www.theatlantic.com/technology/archive/2018/01/equivant-compas-algorithm/550646/

^[viii] https://www.tandfonline.com/doi/pdf/10.1080/13600834.2018.1458455

^[ix] https://www.pib.gov.in/PressReleasePage.aspx?PRID=2178092&reg=3&lang=2

^[x] https://www.thehindu.com/sci-tech/technology/racist-sexist-casteist-is-ai-bad-news-for-india/article67294037.ece

^[xi] https://www.pib.gov.in/PressReleasePage.aspx?PRID=1947490&reg=3&lang=2

^[xii] https://static.pib.gov.in/WriteReadData/specificdocs/documents/2025/feb/doc2025225508901.pdf

^[xiii]https://www.wired.com/story/can-ai-be-fair-judge-court-estonia-thinks-so/#:~:text=All%20products%20featured%20on%20WIRED,percent%20rate%20in%20the%20US.