Solutions to deepfakes, job losses will come from AI itself: Ex IT Secretary

Aruna Sharma served as secretary in the Union Ministry of Electronics and IT and as a member of Reserve Bank of India's high-level committee on deepening digital payments and has played an active role in implementation of new technology.

Artificial Intelligence (AI) has significantly impacted a lot of businesses and is predicted to disrupt the way we work and avail services.

Earlier this year the Government of India came out with the IndiaAI Mission, under which Rs 10,300 crore will be invested to empower startups and expand computer infrastructure access. Aruna Sharma, who served as secretary in the Union Ministry of Electronics and IT and as a member of Reserve Bank of India's high-level committee on deepening digital payments, has played an active role in implementation of new technology and e-governance initiatives in the country.

Speaking to DH’s Gyanendra Keshri, she discusses about AI regulation and the challenges and opportunities associated with the new technology. Edited excerpts:   

How do you see the AI policy evolving in India?

In our regulations for AI, we seem to be confused. It’s not the future we are going to regulate. AI is already with us. Technology moves much faster than the regulator. AI regulation has to be a part of the IT policy. The IT policy should be like an umbrella policy. You have crypto. You have gaming in front of you. Some common factors are going to be there in crypto, gaming, e-commerce etc. In all these areas you are going to see increased use of AI. There has to be provisions to monitor fair play. There should be a code of conduct. There has to be some sort of grievance redressing mechanism and an appellate body. These are common to everybody. AI innovation needs to be facilitated for aggregation, analytics and more and more use cases.

Misuse of personal data through AI: How do you see regulation to prevent that?

As of now we don’t have any AI specific regulation. There are regulations on data storage and data privacy. In the Puttaswamy case, the right to privacy was affirmed. The Supreme Court judgment in the case has established that your data cannot be shared without your permission. This is for the government data, but what about the private companies? Take the example of big tech companies Google and Meta. They are collecting and using a lot of data. Are the data privacy related rules really implanted and monitored in their case, obviously not. Then we are asking the companies to store data in India. But they can take aggregated data offshore for analysis. AI innovation is very quick and growing very fast. The world has already gone ahead with it. We can’t afford to take years and years to come up with a regulation.

What role is the government expected to play in the adoption of AI?

It was the government which took the initiative to bring computerisation in India. The computerisation of the government database created a whole lot of opportunities for the private sector. When computerisation happened, the government took the lead. But in the case of AI, has the government missed the bus? AI is all about managing and using the data. The private sector is using AI for limited things, for example marketing and e-commerce. The biggest database is with the government, in whatever fractured format. There is an immense potential in AI use cases and analytics to promote equity and inclusive growth. This feature has to be harnessed by the government. So, like computerisation, the government will have to take the lead in harnessing the true potential of AI.

What are the key challenges in pushing for the adoption of AI?

When we were pushing for computerisation, the biggest challenge was the division among those who were computer literate and computer illiterate. That kind of a division is going to come. You have a section, which is skilled to handle the AI and there are those who are not skilled to handle the AI. The younger educated ones may equip themselves. However, the younger uneducated ones will be left behind. Even the educated elderly population will be slow in catching up. Then you have the issue of jobs. We have to handle a huge young population who is not appropriately skilled. Jobs in the bigger companies are going to shrink because wherever automation is possible, it will get automated immediately. Hence appropriate skilling and innovation is the key.

How to deal with challenges like deepfakes and potential job losses?

The solutions to the issues like AI-generated deepfakes are also coming from AI. It is being used to detect deepfakes. While there will be job losses, a lot of new jobs are also likely to be created due to AI. Undoubtedly there are going to be a lot of challenges as the use of AI grows, but the solutions are also going to come from the AI itself.