Automated speech recognition (ASR) systems are now used in a variety of applications to convert spoken language to text – from virtual assistants, to closed captioning, to hands-free computing. By analyzing a large corpus of sociolinguistic interviews with white and African American speakers, we demonstrate large racial disparities in the performance of popular commercial ASR systems developed by Amazon, Apple, Google, IBM, and Microsoft. Our results point to hurdles faced by African Americans in using increasingly widespread tools driven by speech recognition technology. More generally, our work illustrates the need to audit emerging machine-learning systems to ensure they are broadly inclusive. See more at fairspeech.stanford.edu.

Speaker:

Allison Koenecke PhD Candidate, Stanford University A fifth-year PhD candidate at Stanford’s Institute for Computational & Mathematical Engineering, Allison Koenecke has research interests that lie broadly at the intersection of economics and computer science, with projects focusing on fairness in algorithmic systems and causal inference in the public health space (full bio).