Directed by Shalini Kantayya, the 2020 documentary Coded Bias details the discovery made by M.I.T. Media Lab Researcher Dr. Joy Buolamwini when she identified a detrimental flaw in facial recognition technology: racial and gender bias. Dr. Buolamwini, a poet of code and founder of the Algorithmic Justice League, has been instrumental in bringing the significant issue to the forefront and speaking up for the need for equitable and bias-free AI.
The documentary unpacks the numerous ways in which machine learning algorithms can be detrimental in exacerbating race, gender, and class-based inequities. The film examines personal stories in cases of employment, housing, and criminal justice, highlighting the real-world, direct impact of bias-ridden technology.
As we’ve seen with facial recognition tools and technologies, machine learning and artificial intelligence are only as just, equitable, and bias-free as their creators. The need for oversight and accountability in emergent technological tools is essential. With funding received from the Sloan Foundation and the Rockefeller Foundation, the Algorithmic Justice League is developing a set of tools that will allow people to report inequities in AI systems.
Dr. Buolamwini’s research reveals that facial recognition systems are often used in ways that are not transparent, leading to potential abuses of power. For example, she discovers that the technology is being used to surveil people without their knowledge, and that it is being used to target certain groups of people. Her research further reveals that there is a pervasive lack of accountability for the biases that are built into facial recognition systems. She uncovered the fact that companies are not held responsible for the potential harm that their technology can cause and that there is no legal framework to protect people from algorithmic bias.
There truly are no safeguards that can guarantee there won’t be an abuse of power of these tools– DR. JOY BUOLAMWINI
