Legal experts urge Centre to come up with specific norms for regulating AI

Last week India took an important step towards regulating artificial intelligence at the Global Partnership on Artificial Intelligence (GPAI) as 29 countries agreed and signed the Delhi Declaration. The declaration highlighted the need to balance the rapid advancements in artificial intelligence systems with addressing concerns related to misinformation, unemployment, transparency, fairness, protection of intellectual property, and personal data.

While the GPAI is just a declaration, experts are hoping that India, too, like the European Union (EU), will come up with specific standards and regulations for artificial intelligence.

This is of significance as the EU is the only region that has taken actual steps towards regulating artificial intelligence. Experts say that while GPAI takes a more non-binding, principles-based approach, leaving implementation details to individual member countries, the EU's AI Regulation provides a much more detailed regulatory structure, outlining specific rules and restrictions for high-risk artificial intelligence applications.

“India should adopt a balanced approach to regulating artificial intelligence going forward by implementing sector-specific guidelines. A one-size-fits-all approach will not work and will end up stifling innovation,” said Anupam Shukla, Pioneer Legal.

“It will also be important to harmonize our approach to artificial intelligence governance with the rest of the world to avoid creating fragmented and conflicting regulations,” he added.

The GPAI ministers affirmed their commitment to promote collaborative artificial intelligence for a global partnership among GPAI members, emphasising equitable access to critical resources for artificial intelligence research and innovation not only to the member countries but beyond, according to the declaration document.

However, Minister of State for Electronics and Information Technology Rajeev Chandrasekhar has said that the government will soon come out with a policy framework to restrict the use of Indian datasets by countries and platforms that are not trusted.

The EU AI Act takes a pioneering role in artificial intelligence regulation, categorizing risks into unacceptable, high-risk, generative, and limited-risk. It bans threats like cognitive manipulation and facial recognition, assesses high-risk artificial intelligence systems before market entry, requires transparency for Generative AI (e.g., ChatGPT), and sets minimal transparency standards for limited-risk artificial intelligence.

The Act also has stringent provisions for violations that may incur penalties up to €30 million or 6% of the violating entity's global income.

“The Government should devise policy to support such innovation with global partnership and resource sharing. Regulation of artificial intelligence is as important as its promotion,” says Astha Sharma, Partner, AQUILAW.

She further says that India can take the example of the EU AI Regulation to enact similar laws to regulate artificial intelligence systems in India.

The EU Act is in a sense, the torchbearer for artificial intelligence regulation, whereas the New Delhi GPAI declaration and the Bletchley declaration aim more towards creating awareness around artificial intelligence regulation, safety, and innovation, says Sharma.

“Even though the declaration is a step forward in artificial intelligence innovation, there is little said about the risks posed by artificial intelligence, such as privacy, safety, bias, etc. The declaration is also silent on the policy measures to mitigate such risks,” she comments on the New Delhi declaration.

As India aspires to lead in artificial intelligence innovation, legal experts advocate a nuanced, sector-specific regulatory strategy. The GPAI New Delhi declaration highlights global collaboration, while insights from the EU's AI Regulation offer perspectives for India's evolving approach, emphasising the imperative of balancing innovation with robust regulation.

The need to balance regulation and innovation should be the focus of India’s artificial intelligence strategy and the EU AI Act can be a good reference point to look at, experts feel.