AI and Racism

The unprecedented rise of AI in almost every sector of society has its benefits, but what about its downfalls? AI is being used in nearly every institution for a variety of reasons. Companies are using it to improve workplace efficiency, marketing metrics, and the hiring process. Universities and professors are using AI to create lesson plans, detect the use of AI by students, and aid in sifting through data used in research. The government is using AI through the use of Flock cameras to detect faces and cars to scan for and prevent criminal activity. All of this can be beneficial in the grand scheme of things, but when we look closely at AI, there are extreme pitfalls. 

AI itself was created by humans through coding and development. This opens up doors for subconscious and conscious biases to appear in AI-generated responses. We often view AI as unbiased because it is not human and uses the internet to answer questions. Research has shown that AI can produce responses that perpetuate racial discrimination because LLMs use historical inequities and biased data to train themselves. This is reflected in a number of ways. For example, AI screening tools used in the hiring process are more likely to reject minorities based on their names, relevant work experience, and educational background. Facial recognition software has been shown to produce higher error rates for individuals with darker skin tones. This becomes a significant problem when taking these outcomes and viewing them as unbiased. Not thinking critically about how AI operates can further perpetuate historical and current inequalities in society. 

Not only does AI produce discriminatory responses and operations, but data centers that store the brain of large LLMs reflect issues surrounding environmental racism. Big tech companies will often target towns that are low-income and have higher non-white populations to build data centers. This reflects historical patterns of resource exploitation and redlining used to discriminate against minorities. Individuals living in these towns often lack the political capacity and power to resist against million-dollar companies. Tech companies offer big buyouts to local politicians that they simply can not refuse. Communities are forced to bear the burden of air and water pollution as well as the economic impact that comes with the upkeep of data centers. 

Honest and deep conversations are needed to combat the discrimination and environmental racism that comes with the expansion of AI in everyday life. Communities need to set differences aside to fight back against tech companies. There also needs to be a general agreement and understanding that AI is biased and not a fully reliable source.