The Potential for Discrimination in AI Systems
Artificial intelligence (AI) is increasingly being used to make important decisions that can have a significant impact on people’s lives. Whether it’s college admissions, mortgage loan applications, job interviews, or parole decisions, AI systems are now playing a major role in these processes. However, there is a growing concern that these systems can amplify existing biases and perpetuate discrimination. Researchers at Morgan State University have identified the potential for racial, gender, and other forms of bias to be magnified by AI systems. Gabriella Waters, the director of the Center for Equitable Artificial Intelligence and Machine Learning Systems (CEAMLS) at Morgan State, warns that if biases are embedded in these systems, they can lead to significant and widespread discrimination. Waters emphasizes that it is the responsibility of developers and researchers to ensure that AI systems do not introduce or perpetuate bias.
Morgan State University’s Center for Equitable AI
Morgan State University’s CEAMLS was established in July 2022 with the aim of addressing the inherent problems in existing AI systems and developing technologies that are free from bias. The center brings together experts from various disciplines, including engineering and philosophy, to harness the power of AI while ensuring fairness and equity. As a historically Black university, Morgan State has a unique position in the field of AI. Kofi Nyarko, the director of CEAMLS and a professor of electrical and computer engineering, believes that the university can contribute not only to advancing the science of AI but also to ensuring that the community is involved in the process. With significant funding from both public and private sources, Morgan State is investing in AI research to make a lasting impact.
Addressing Bias in AI Systems
The inherent bias in AI systems is not the fault of the machines themselves but rather a reflection of the data they are trained on. Humans are the source of the data sets, and their biases and prejudices can unwittingly seep into these systems. It is crucial to recognize that bias is not always intentional but often a result of human nature. Data can sometimes obscure the truth and lead to erroneous conclusions. For example, correlations between high ice cream sales and shark attacks in certain areas do not imply causation. Therefore, addressing bias in AI systems requires a multidisciplinary approach that examines the data, algorithms, and decision-making processes to ensure fairness and accuracy.
Maryland’s Efforts to Harness the Power of AI
Maryland’s General Assembly and the administration of Governor Wes Moore are taking active steps to explore the potential of AI and its responsible use in state government services. Their goal is to harness the power of AI while upholding core values such as equity, security, and privacy. The state is striving to understand the applications of AI and ensure that it aligns with the principles of fairness and non-discrimination. Nishant Shah, Moore’s senior advisor for responsible AI, is spearheading the development of AI principles and values that will guide the procurement and monitoring of AI systems. Efforts are also underway to increase the knowledge and skills of state personnel in AI to effectively leverage its potential.
The Importance of Ethical Capacity in AI Innovation
Morgan State’s CEAMLS recognizes the need for thoughtful deliberation and reflection when it comes to the development and implementation of socially disruptive technologies like AI. J. Phillip Honenberger, a philosophy and religious studies expert at the center, believes that AI innovations should not be stifled due to the fear of unintended consequences. Instead, he emphasizes the importance of building ethical and reflective capacity to ensure that AI is developed and used responsibly. Looking back at the impact of social media on politics, Honenberger highlights the lack of foresight and contemplation that often accompanies technological advancements. The center at Morgan State aims to fill this gap by promoting ethical decision-making and conscious development of AI systems.
As AI becomes increasingly integrated into various aspects of society, it is essential to address the potential biases and discrimination that can arise. Morgan State University’s CEAMLS is at the forefront of this effort, working to ensure that AI systems are fair, equitable, and unbiased. Maryland is also taking proactive measures to harness the power of AI while upholding core values. With a focus on ethical capacity and responsible innovation, the goal is to leverage AI for the benefit of society while minimizing any unintended negative consequences.
Analyst comment
Positive news: The Potential for Discrimination in AI Systems
As an analyst, it is expected that there will be a significant focus on addressing potential biases and discrimination in AI systems. Morgan State University’s CEAMLS is leading this effort, working to develop AI technologies that are free from bias. Maryland’s government is also actively exploring the responsible use of AI in state services. By promoting ethical decision-making and responsible innovation, the goal is to leverage AI for societal benefit while minimizing unintended negative consequences. The market for AI systems and research is likely to grow as demand increases for fair and unbiased AI technologies.
