The artificial intelligence (AI) ecosystem must be free of bias, and facial recognition technology cannot exhibit bias on the basis of colour or ethnicity, Ravi Shankar Prasad, Union Minister for Electronics and Information Technology said. Prasad also prescribed a set of principles, or standards, that are necessary for the empowerment of Indian citizens and the development of the AI ecosystem in the country.
Prasad was speaking at the event all.ai, organised by Intel. The four-day event, which was inaugurated on Monday, will feature discussions on the applications of artificial intelligence in various areas such as healthcare, education and governance. Telangana IT Secretary Jayesh Ranjan inaugurated an applied AI research centre INAI, a partnership between Intel, IIIT-Hyderabad, PHFI and the Government of Telangana.
Prasad noted the need for transparency in algorithms. He argued that users of an algorithm should have all the necessary information at hand. He also argued for the citizens’ right to privacy, and how data must not be collected without their consent.
- Free of bias: Prasad said that the internet is a global platform that had truly brought meaningful change in the citizens’ lives. However, this global platform must not discount its local connect. “The global and local must be in sync,” he said. The minister was arguing that while AI had a great potential for applications in facial recognition technology, it had to take into account the data of local citizens. “The facial recognition process cannot show any bias of colour or ethnicity,” he said .
- Trustworthy data sharing: “Data is the oxygen for AI. The data ecosystem has to be trustworthy,” said Prasad. He argued that users needed to be informed the purpose for which their data is being used. “There must be a sanctity in that process, be it safety, security or consent,” he said.
- Privacy has to be respected: Prasad noted that as citizens of a democracy, the rights of ordinary Indians were paramount. “It is equally important that the right to privacy is respected as well,” he said. The minister said that privacy would be respected only if the data collected from a citizen is based on consent — it is taken voluntarily.
- Social empowerment the key goal: The minister argued that whoever the practitioner of an AI-based tool may be — government of private — its ultimate goal needs to the improvement of the lives of citizens. These issues may be related to healthcare, or the empowerment of farmers, he said.
- AI can’t be weaponised: “The application of AI should not be in a manner that the basic time-tested attributes of human existence are lost. As Prime Minister Narendra Modi has pointed out, AI cannot be used a means of weaponisation,” said Prasad.
The IT minister also commented on the legal aspects surrounding the use of AI. He said that Personal Data Protection (PDP) Bill, which is under consideration of a joint Parliamentary Committee, would provide a robust law, that will empower the development of AI in India. He noted that there are other issues that will be tackled with as the time comes. He tool the example of self-driven cars: “As an IT minister, the prospect excites me. But as the law minister, it concerns me. If a driverless car causes an accident, who is responsible: the car, or its owner, who might not even be in it?”
Prasad also addressed concerns about the loss of jobs because of AI. “Yes, there are apprehensions that AI will displace jobs. But if some jobs will go, more and more will be created. AI may displace existing jobs, but more jobs will be created once the ecosystem unfolds.”
