Sanas, a Silicon Valley startup, developed AI software that removes or alters accents in real-time during call center conversations, sparking controversy over accusations of promoting racial bias by making workers sound 'white' rather than addressing underlying prejudices.
Sanas, founded by three former Stanford students and based in Palo Alto, has developed AI software that removes or alters accents in real-time during call center conversations. Since its August 2021 launch, the company has raised over $32 million in venture funding, including a $32 million Series A round in June 2022. The technology transforms voices with non-American accents into what the company calls 'standard American English' or what critics describe as 'white-sounding' voices. At least seven outsourcing firms have already deployed Sanas products in their call centers. The software has sparked significant controversy, with critics arguing it reinforces racial bias rather than addressing it. Experts note that the technology forces workers, particularly those in countries like India and the Philippines, to further mask their identities rather than challenging caller prejudices. The company claims the software helps call center workers avoid harassment and misunderstandings, but researchers argue it perpetuates existing power imbalances and fails to address the root causes of bias. The technology operates by having call center companies maintain administrative control, effectively removing worker autonomy over its use despite company claims about worker choice.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Unequal treatment of individuals or groups by AI, often based on race, gender, or other sensitive characteristics, resulting in unfair outcomes and unfair representation of those groups.
AI system
Due to a decision or action made by an AI system
Intentional
Due to an expected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed