With “Coded Bias,” Shalini Kantayya wanted to explore “the dark underbelly of big tech.” She was following the lead of Joy Buolamwini, an MIT media lab researcher who discovered that facial recognition does not see dark-skinned faces accurately.

Initially drawn to computer science because it seemed “detached from the problems of the real world,” Buolamwini quickly realized that it’s human beings — ones who are predominantly male and lighter-skinned — who teach AI how to see, and issues of bias inevitably creep into technology. When people offer machines skewed data, machines yield skewed results: facial recognition software is often inaccurate, and people get misidentified, directly impacting access to jobs and healthcare, as well as threatening their safety.

Kantayya’s doc makes a convincing — and rather terrifying — case for why blind faith in data isn’t just misguided, but dangerous. Tech is evolving at a rapid pace, and we need safeguards in place to ensure that racism, sexism, and other forms of prejudice aren’t being mechanized. By providing an overview of the situation and digging into particular case studies, “Coded Bias” serves as a wakeup call as well as a call to arms. It follows Buolamwini and others’ push for legislation in the U.S. to govern against bias in algorithms that affect so many facets of our lives.

“I want people to ask critical questions about so many of the technologies that we interact with every day,” Kantayya told us.

“Coded Bias” isn’t just directed by and centered around a woman. The doc is dominated by women. Expert interviews tend to skew pale and male, especially in STEM docs, but “Coded Bias” highlights the voices of women, and particularly women of color, in a rare and welcome change.

“Coded Bias” is now virtually screening at New York City’s Metrograph and opens in additional cities November 18.