Silicon Valley start-up using tech to make Indian voices ‘sound whiter’
The firm is accused of ‘bias and racism’. Representational image (iStock)
A Silicon Valley start-up has been converting Indian-accented voices using Artificial intelligence to make them sound American, according to media reports.
The start-up Sanas has told the BBC that its technology could overcome accent-based prejudice and reduce racist abuse faced by staff.
But, the firm has been accused of ‘bias and racism’ for using AI e to make accents ‘sound whiter’. Critics said that it is a service in the wrong direction as language diversity should be celebrated.
According to the company, the accent-altering technology can lead to more polite interactions between callers and customer service representatives.
“We don’t want to say that accents are a problem because you have one. They’re only a problem because they cause bias and they cause misunderstandings,” Marty Sarim, Sanas president, was quoted as saying in media reports.
He claimed that the company’s technology will bring millions of jobs to the Philippines, India and other countries.
Sharath Keshava Narayana, a co-founder of Sanas, said that all four of the founders were immigrants and so were 90 per cent of the start-up’s employees.
Narayana, a former call centre agent himself, said the tool had been inspired in part by the experience of a close friend of one of the other founders. The company believes its technology can prevent abuse, he added. Sanas received $32m in funding since June 2022.
Nayana Prakash, a researcher at the Oxford Internet Institute, said the technology “reinforces and permits the biases people have towards Indian accents.
“Outsourced workers in India have suffered from this kind of bias and racism for decades – in many ways this technology is just a glamourised solution for something that has been going on for a long time,” she was quoted as saying by The Telegraph.
“The idea that AI should be used as a solution when the problem is the underlying bias and anxiety of Western consumers really places the issue squarely on the shoulders of the workers, rather than questioning these prejudices as they occur in the customers.”
Ashleigh Ainsley, co-founder of the organisation Color in Tech, said: “Should we just alter people’s skin colours because some people might not like that because they’re racist?
“We cannot move in this direction. We need to build tolerance.”
Shalu Yadav, a BBC journalist based in Delhi who worked at three call centres to earn extra money as a student, said employers expected her to learn about American culture and use an American accent.
A Aneesh, a sociologist and the incoming director of the University of Oregon’s School of Global Studies and Languages, told the Guardian that workers had to relearn pronunciations to stay on the call centre job.
“Workers had to relearn pronunciations of words such as “laboratory”, which Indians pronounce with the British stress on the second syllable. They also had to eliminate parts of Indian English – like the frequent use of the word “sir”,” Aneesh, who has spent years studying call centers and accent neutralization, said.
“They had to learn uniquely American words, including a list of over 30 street designations such as “boulevard”, and memorize all 50 US states and capitals. They have to mimic the culture as well as neutralize their own culture.”
Sanas revealed that firms were testing their technology for internal use, to ease communication between teams in Korea and the US or between teams in North and South India.