Microsoft warns of danger posed by mass surveillance unless facial recognition is regulated
The future is with us, now. Computers have been taught to identify the human face.
The technology is used by police to pick out suspects in a crowd and by banks to authenticate their customers. But what if the computer gets it wrong?
Microsoft is at the vanguard of facial recognition. The company’s president, Brad Smith, says the technology has the potential to solve many of society’s problems but also has dangerous limitations.
“One challenges with the technology in its current form is it doesn’t work as well for women as it does for men, it doesn’t work as well for people of colour,” Smith told ITV News.
“It’s more likely to find error, it’s more likely to find a mismatch, it more likely to fail to identity you, it’s more likely to identify as someone you not."
Microsoft says facial recognition technology not only has weaknesses but is open to abuse and needs tighter regulation by governments.
In a world where computers and cameras are everywhere, there are implications for mass surveillance.
“If you enable a government to just follow somebody our for example anybody all the time then ability of people to assemble top speak the fundamental cornerstones of democracies can be undermined.”
Brad Smith says Microsoft has rejected business in countries which it felt wanted to use facial recognition in a way that would put human rights at risk and the company has turned down deals with law enforcement agencies which it felt might use the technology in a way that risked bias or discrimination.
“That was the right think for Microsoft to but it’s not healthy to create a market where people are forced to choose between doing right and gaining business. A floor is needed, that’s what regulation can do”.
A government spokesperson said: "The use of AI in technology like facial recognition has vast possibilities but it's crucial that the public have confidence in it being used responsibly.
"That's why we've set up the world's first Centre for Data Ethics and Innovation, to shape how data and AI are used, ensuring they deliver maximum benefits for society."