This algorithm will make call centre employees sound 'American'

Privacy and surveillance researcher Chris Gilliard told The Guardian that a step like this leads to the "erasure of people as individuals"

By
Web Desk
Woman in call centre. — Unsplash
Woman in call centre. — Unsplash

  • American businesses outsource call centre work to other countries.
  • Sanas develops software that can alter users' voices and "Westernise" them. 
  • While founders believe they are empowering people, others do not agree. 


It is now common knowledge that American businesses outsource their call centre work to countries in the developing world like Pakistan, the Philippines, and India to cut costs and protect themselves from the States' labour laws.

Telecommunication jobs can be very draining with low pay, difficult hours, and often abusive consumer interactions.

In an effort to change how domestic consumers interact with call centre employees, a startup is launching an AI that will make employees sound "American" so that it boosts non-native English speakers' well-being.

However, critics believe it could be counterproductive.

Sanas is the company that is working since 2020 to develop AI software that can alter users' voices and "Westernise" them. 

"Sanas’ AI engine can transform a speaker’s accent into what passes for another one,” explained The Guardian.

While the software can improve human communication, the psychological and ethical implications are clear.

It shows that the deep-rooted social issues are not being dealt with and the individuality of non-native speakers is being devalued, reported Popular Science.

One of the co-founders has worked in Indian call centres which are probably why the company claims on its website that it aims to "deepen empathy". 

Co-founder Sharath Keshava Narayana claims that more than 1,000 call centres in India are already using their product.

While the founders believe they are empowering people, others do not agree. 

Privacy and surveillance researcher Chris Gilliard told The Guardian that a step like this leads to the "erasure of people as individuals".

He said that such software would homogenise everything and neglect the "beauty" that comes with the diversity of voice and accents.