India’s AI Safety Institute should strike balance between ‘prevention’ and ‘promotion’, says Raghunath Mashelkar

India does not have any specific AI-related regulatory arrangement.

India should consider establishing an AI Safety Institute on the lines of those existing in the UK and the US to ensure that increasing use of sophisticated artificial intelligence technologies can benefit everybody in the world’s most populous nation while simultaneously minimizing risks associated with the tech.

Such is the view of Dr Raghunath Mashelkar, one of India’s foremost scientists, and former director-general of the Council of Scientific & Industrial Research (CSIR).

“We need an AI Safety Institute here and now,” Mashelkar said during a freewheeling conversation with this writer on artificial intelligence. Given the breakneck speed at which AI technologies are being developed, an AI Safety Institute in India could ensure that the benefits of artificial intelligence can be maximized, and ‘responsible AI’ getting a boost, he added.

AI Safety Institute’s Own Distinct Identity

However, Mashelkar pointed out that an AI Safety Institute in India – while taking inspiration from similar set-ups elsewhere – does not need to be an exact replica of what is there in other countries but aim to have its own distinct identity. Such a body should also include representatives from industry, he added.

A part of the United Kingdom’s Department for Science, Innovation and Technology, the mission of the AI Safety Institute in the United Kingdom “is to minimise surprise to the UK and humanity from rapid and unexpected advances in AI”. Housed within the National Institute of Standards and Technology (NIST) of the United States, the US AI Safety Institute aims to advance “the science, practice, and adoption of AI safety across the spectrum of risks, including those to national security, public safety, and individual rights.”

The Bletchley Declaration – that was agreed on at the AI Safety Summit held in the UK in November 2023 where India was also a participant – had taken note of multiple risks posed by artificial intelligence, including its capability to “manipulate content or generate deceptive content”.

AI Regulation

Mashelkar – also a former chairman of the National Innovation Foundation – said that as is being done by other nations, India, too, should put in place its own “comprehensive AI regulation framework” to govern the use of artificial intelligence in the country. Such a framework in India should strike a fine balance between “prevention”, “permissiveness”, and “promotion” so that artificial intelligence could flourish here, and unintended consequences of the tech could be lessened.

As of now, India does not have any specific AI-related regulatory arrangement.

Incidentally, on March 21, the United Nations General Assembly adopted a landmark resolution on the promotion of “safe, secure ad trustworthy” AI systems. It is the first time that the Assembly has adopted a resolution on regulating artificial intelligence.

AI & jobs

Mashelkar said a disruption in the jobs landscape in India due to rising use of AI technologies can initially be expected in the country in line with what has been forecast for the rest of the world. SkillsFuture Singapor is a good model that India could look at emulating in this connection, he added.

Mashelkar said India may consider providing social security to people whose job losses may be directly attributable to the deployment of artificial intelligence at their places of work. “You can have some form of a universal basic income (UBI) for them”, he pointed out. Such a measure, though, should be for a “fixed period of time” covering only the period till these individuals are upskilled to find new employment, he added.

On January 17, IMF’s first deputy managing director Gita Gopinath had, in a video posted on the social media platform X, highlighted that 40 percent of jobs worldwide are “exposed to AI”, with that figure rising to 60 percent in the case of the United States and dipping to 26 percent in “low-income countries”. She had said then that while this AI “exposure” could raise productivity, it could also lead to “displacement” of workers.

Gopinath had further said that it was “important for countries to make sure that their labour market policies and their tax policies are consistent with this technology being on net (a) benefit to humans”.

AI expertise in government

Mashelkar said that thought could be given towards increasing the pool of government employees having expertise in the AI arena at both the Centre and at the level of states primarily through upskilling measures. Having more people with the knowledge and understanding of how AI works could further improve the quality of data-driven governance and delivery of various citizen-centric services.

The added benefit could come in the form of authorities being able to come up with effective safeguards to ensure that AI technologies are used responsibly, and misuse of the tech can be prevented, he pointed out.

AI & innovation

Mashelkar said India has the talent pool that is required to redefine the world of artificial intelligence. Some of the measures that could enable India to make a bigger mark in the AI domain are a “long-term vision and commitment”, “collaborative partnerships”, R&D being incentivized more, coming up with technologies that could be commercialised, and, also, the country’s industry leaders playing a more active role in driving the international discourse on artificial intelligence, he added.

Incidentally, Reliance Industries chairman Mukesh Ambani had earlier flagged the infrastructure challenge faced by India on the AI front. Addressing shareholders at the company’s 46^th annual general meeting held on August 28, 2023, Ambani had said: “India has the scale. India has the data. India has the talent. But we also need digital infrastructure in India that can handle AI’s immense computational demands.”