Accountability in Artificial Intelligence | American Civil Liberties Union

Question 1

Why is Artificial Intelligence (AI) biased?

Accepted Answer

AI is built by humans and deployed in systems and institutions that have been marked by entrenched discrimination — from the criminal legal system, to housing, to the workplace, to our financial systems. Bias is often baked into the outcomes the AI is asked to predict. Likewise, bias is in the data used to train the AI — data that is often discriminatory or unrepresentative for people of color, women, or other marginalized groups — and can rear its head throughout the AI’s design, development, implementation, and use. The tech industry’s lack of representation of people who understand and can work to address the potential harms of these technologies only exacerbates this problem.

Question 2

How does AI discriminate in housing?

Accepted Answer

AI systems used to evaluate potential tenants rely on court records and other datasets that have their own built-in biases that reflect systemic racism, sexism, and ableism, and are notoriously full of errors. People are regularly denied housing, despite their ability to pay rent, because tenant screening algorithms deem them ineligible or unworthy. These algorithms use data such as eviction and criminal histories, which reflect longstanding disparities in housing and the criminal legal system. People of color seeking loans to purchase homes or refinance have also been overcharged by millions thanks to AI tools used by lenders, and automated valuation models trained on discriminatory data further exacerbate longstanding undervaluations in home appraisals in communities of color.

Question 3

How does AI discriminate in hiring?

Accepted Answer

Many employers now use AI-driven tools at virtually every stage of the hiring process, from attracting applicants to screening and interviewing job seekers, and these tools pose enormous risks for discrimination against people with disabilities, people of color, and other protected groups. Rather than help eliminate discriminatory practices, AI has worsened them — hampering the economic security of marginalized groups that have long dealt with systemic discrimination.

News & Commentary about Accountability in Artificial Intelligence