Advocating for a New Profession: The AI Bias Investigator
Nsikak Essien - M.
Finance Professional at Holcim | Holder of MSc in Economics, Accounting, and Supply Chain with expertise in Finance & Risk. Designed and implemented transformation roadmaps, leading to successful finance transformation
In an era where artificial intelligence (AI) systems are increasingly woven into the fabric of daily life, from hiring decisions to criminal justice, the need for accountability and fairness in these systems has never been more critical. AI, while powerful, is not infallible. It can perpetuate and even exacerbate biases, particularly against marginalized communities, including Black individuals. To address these issues, I advocate for the establishment of a new profession: the AI Bias Investigator. This role would be essential in identifying, understanding, and mitigating biases in AI systems, ensuring that these technologies serve all people equitably.
The Urgency of Addressing AI Bias
AI systems learn from data, and if that data reflects societal biases, the AI can replicate and even amplify those biases. Numerous studies have highlighted how AI systems, particularly those used in facial recognition, hiring, and law enforcement, disproportionately impact Black communities. For instance, a 2019 study by the National Institute of Standards and Technology (NIST) found that facial recognition algorithms were up to 100 times more likely to misidentify Black and Asian faces compared to white faces (Grother, Ngan, & Hanaoka, 2019). This discrepancy is not just a technical flaw—it can lead to wrongful arrests, denied opportunities, and perpetuation of systemic racism.
One notorious example is the case of Robert Williams, a Black man who was wrongfully arrested in Detroit in 2020 based on a faulty facial recognition match (Hill, 2020). The AI system used by the police incorrectly identified Williams as the suspect in a robbery, leading to his arrest despite clear evidence of his innocence. This case, among others, underscores the critical need for professionals who can rigorously investigate and address such biases in AI systems.
The Role of an AI Bias Investigator
An AI Bias Investigator would be a specialized professional responsible for scrutinizing AI systems to identify and mitigate biases. This role would combine expertise in data science, ethics, and social justice to ensure AI technologies are fair and just. The responsibilities would include:
Emphasizing Black Discrimination in AI
The creation of the AI Bias Investigator role is particularly urgent when considering the discriminatory impact AI has had on Black communities. AI systems are often trained on datasets that underrepresent Black individuals or reflect existing racial prejudices. This can result in biased outcomes in areas like hiring, lending, and law enforcement.
In hiring, for example, AI-driven tools used by companies to screen resumes have been found to discriminate against applicants from minority backgrounds. Amazon, for instance, had to scrap its AI recruiting tool after it was found to be biased against women and minority candidates (Dastin, 2018). Similarly, AI systems used in lending have been shown to offer less favorable terms to Black applicants compared to white applicants, perpetuating economic disparities (Bartlett, Morse, Stanton, & Wallace, 2020).
Decided Cases and Examples
One landmark case that highlights the dangers of AI bias is the case of State v. Loomis (2016), where the defendant, Eric Loomis, challenged the use of COMPAS, an AI-driven risk assessment tool, in his sentencing. Loomis argued that the system was biased against him as a Black man and that its proprietary nature made it impossible to challenge its accuracy or fairness. While the court upheld the use of COMPAS, the case sparked widespread debate about the fairness and transparency of AI in the criminal justice system (Angwin, Larson, Mattu, & Kirchner, 2016).
Moreover, in the 2019 case involving the wrongful arrest of Robert Williams, the failure of the AI system to accurately identify him based on race was not just a technical error—it was a stark reminder of how AI can reinforce racial discrimination. These cases underscore the need for AI Bias Investigators who can independently review such systems and advocate for changes that protect individuals from biased outcomes.
Conclusion
As AI continues to play a more prominent role in critical aspects of society, ensuring that these systems operate fairly and justly is imperative. The establishment of the AI Bias Investigator profession would be a significant step toward addressing the systemic biases that plague AI technologies, particularly those that disproportionately harm Black communities. By creating this role, we can help ensure that AI serves as a tool for progress and equity, rather than perpetuating the inequalities of the past.
References: