10.46243/jst.2025.v10.i01.pp26-38 registered

The focus of this paper is an examination of the ethical dilemmas and biases present in decision-makingsystems within the field of AI focusing on their implications for fairness, transparency, and accountability. Justas AI becomes imbued into various sectors such as health, finance, policing, or hiring, the concern for thepotential bias in outcomes and discrimination has gained much more traction. The historical data that underliethe training of the AI systems are essentially reflective of the social biases that exist in society now. These misusesmake a cycle out of the discrimination and augmenting it. Mostly, we must understand how biases pallor explicitor itself impute in AI algorithms and consequences to the communities suffering the most, that is, race, gender,and socio-economic status. The paper also enquires into the ethical dilemmas resulting from the applicationpossibility of inflicting non-intentional harm. It highlights additional issues surrounding AI transparencybecause many of them are "black boxes" regarding which understanding how the decisions were made isdifficult. Some recommendations for potential solutions are also provided, which include the need for moreregulation, creating more diverse data sets, implementing greater transparency in algorithms, and bringing ininterdisciplinary teams during the AI development process to curb bias and improve fairness. Addressing theseethical challenges will secure the successful application of AI systems for society in a fair way in a way that doesnot encourage harmful bias or discrimination.