Global AI Conclave: India may have balanced view on AI regulations focussed on citizens and markets, say panellists

Sindhu Gangadharan, MD, SAP Labs India and Mary Snapp, vice president- strategic initiatives Microsoft, at CNBC-TV18 & Moneycontrol’s Global AI Conclave.

By combining the EU Artificial Intelligence (AI) Act and the US executive order on regulating AI, India can have the best of both worlds by focussing on the citizens and small and medium enterprises (SMEs) and as well as regulating the markets, said panellists Sindhu Gangadharan, MD, SAP Labs India and Mary Snapp, vice president- strategic initiatives Microsoft, at CNBC-TV18 & Moneycontrol’s Global AI Conclave.

This comes at a time when the Indian government is evaluating possible AI regulations and the areas that require tighter regulations.

Gangadharan said, “EU Act is more focussed on citizens while the American executive order is focused on the markets as well. And in India, I think it needs to be a balanced hybrid of these two. We have very specific needs catering to certain industries and segments.”

“If you look at the EU Act for AI, that’s the most comprehensive regulations on AI we have. It’s really about not having extremely rigid rules in place, but if the use cases are high risk, of course, you can’t deploy those, but if it’s moderate and low risk, it can be deployed after giving transparency to the end users,” she added.

Snapp said, “I am impressed by the focus here on small and medium businesses and bringing out regulations surrounding that will also make India stand out.”

She added, “We have been strong proponents of regulations. We believe regulations can spark innovation in certain cases while ensuring safety. It’s really important that we look at standards and existing frameworks in India and the US, it’s important to go granular sector wise and work with academia and non-profits to build regulations.”

Speaking on a possibility of having a common global framework around AI and China being involved, Snapp said, “I think we are on the path to do that. We have heard the Prime Minister’s (Narendra Modi) point of view on China. I believe responsible countries need to move forward no matter what and need to continue to bring others in. We think we are getting aligned and…because there is an outlier doesn’t mean we need to put our heads in the sand.”

Internal AI committees

Interestingly, both the companies spoke about having dedicated internal AI committees of expert years before generative AI became a hype.

Snapp shared that around 6-7 years back, Microsoft had created a committee including highest levels of company executives to review the use cases of AI. Then four years ago, they created an office within the legal organisation and deployed dedicated “AI champions” throughout engineering teams to identify impact of the use cases of AI in our products.

“The committees identity risks and try to mitigate them. In the situation of sensitive use cases -- if it could cause harm the legal rights of individuals, physical and psychological issues and human rights violations, it would be taken down,” she said.

Gangadharan highlighted that about 87 percent of the world’s business transactions go through SAP’s system. Around 99 of the top companies of the world run on SAP. Nearly 8 out of 10 car companies running in India, are SAP’s customers; which means, SAP has to be responsible.

“Every AI process be it cash, sourcing, design processes goes through a very well defined AI ethics committee, which has been around for several years now. Data transparency requires human being in the loop depending on specific use cases and industries,” she said.

Snapp later even got candid about the recent fiasco at Open AI, related to Sam Altman’s ouster as the CEO and rehiring and what went behind the scenes at Microsoft.