Algorithmic Discrimination: Continuation of Human Bias or a Gateway to Equality?
Artificial Intelligence (AI) has rapidly emerged as a cornerstone of modern technology, profoundly influencing various aspects of daily life. From determining creditworthiness to making critical decisions in healthcare, AI systems are now at the forefront of decision-making processes. However, as these systems become more integral to societal functions, questions arise regarding their fairness and potential for discrimination. The pivotal question remains: does AI simply mirror human biases, or can it be a tool to foster equality?
The Risks of Algorithmic Discrimination
At the heart of the debate is the concept of algorithmic discrimination. This occurs when AI systems, which rely heavily on historical data, accidentally reinforce the biases present in that data. For instance, AI systems used in the judicial systems have been shown to exhibit racial biases, with black defendants often labeled as high-risk for reoffending compared to white defendants under similar circumstances. Similarly, gender biases have surfaced in financial systems, such as the Apple Card controversy where women were provided lower credit limits compared to men.
The challenge lies in the fact that AI, unlike humans, does not possess inherent ethical or moral judgments. It processes data objectively, meaning any bias present in the input data is likely to be reflected in the output. This has led to numerous cases where AI systems have made discriminatory decisions, further encouraging societal biases
Legal Frameworks Addressing AI Discrimination
The regulation of AI to prevent such discriminatory practices is a pressing issue. While current international treaties addressing discrimination, such as the Universal Declaration of Human Rights and the International Covenant on Civil and Political Rights, do not explicitly cover AI, they lay the groundwork for emerging regulations.
The European Union's AI Act and the Council of Europe's Framework Convention on Artificial Intelligence and Human Rights, Democracy, and the Rule of Law are pioneering efforts in this regard. These legal instruments aim to ensure that AI systems operate within a framework that prioritizes human rights, non-discrimination, and transparency. They represent a crucial step toward regulating AI at a fundamental level, addressing both the potential benefits and risks associated with these technologies.
The Role of Internal Policies in AI Companies
Beyond international regulations, the internal policies of AI companies play a critical role in mitigating algorithmic discrimination. Companies like OpenAI have established usage policies that prohibit discriminatory practices, violence, and illegal activities. These policies are enforced through a combination of automated systems and human oversight, ensuring that AI systems adhere to ethical standards.
However, the effectiveness of these internal policies depends on the transparency and accountability of the companies. For AI to truly serve as a gateway to equality, it is essential that companies not only implement robust ethical guidelines but also ensure that their AI systems are regularly audited for compliance
What Needs to be Regulated?
Regulating AI to prevent discrimination involves two key aspects: the quality of the data used to train algorithms and the transparency of the decision-making process. Ensuring that AI systems are trained on unbiased, high-quality data is crucial to preventing the perpetuation of existing biases. Moreover, transparency in how AI decisions are made allows for easier identification and correction of discriminatory practices.
Transparency and accountability are connected principles that are critical to the ethical operation of AI systems. Companies and regulatory bodies must work together to ensure that AI systems are not only efficient and innovative but also fair and just.
Conclusion
AI represents a significant technological leap with the potential to either foster or alleviate societal inequalities. The key to ensuring that AI serves as a force for good lies in robust regulation, ethical business practices, and a commitment to transparency. As we move forward, it is imperative that legal frameworks evolve to address the unique challenges posed by AI, ensuring that this powerful technology is used to promote equality and justice in society.
By addressing the potential for algorithmic discrimination through comprehensive legal regulations and ethical standards, we can harness the power of AI to create a more equitable future.
* This blog is the result of a research project authored by Jovan Jablan (Legal Consultant at SLC), Luka Laković (Teaching Assistant at the Faculty of Information Systems and Technologies at the University of Donja Gorica), and Andrea Mićanović (Legal Consultant at SLC).
