Tech companies say no to more laws, statutory body to regulate AI

This comes after the Telecom Regulatory Authority of India (TRAI) released a 138-page report with recommendation on ‘Leveraging Artificial Intelligence and Big Data in Telecommunication Sector’ on Thursday.

TRAI has suggested that an independent statutory authority, ‘Artificial Intelligence and Data Authority of India (AIDAI),’ be established immediately for development of responsible AI and regulation of use cases in the country.

“The development and deployment of AI needs oversight to ensure responsible use and scale up of adoption, but we strongly recommend against creating more laws or statutory authority to regulate AI,” Ashish Aggarwal, vice president and head of public policy at Nasscom, told ET. A disjointed and disproportionate regulatory intervention can impede the growth of the evolving AI ecosystem in India and cause avoidable uncertainty, he argued.

Nasscom comprises over 3,000 companies from the technology industry.

Strengthen existing framework

Discover the stories of your interest

The existing regulatory framework and enforcement capacity should be updated to deal with the deployment of AI happening at large scale, Aggarwal said. “We had submitted the same to TRAI to ensure that they do not replicate existing efforts,” he pointed out.

TRAI has suggested that its previously recommended statutory body, Data Digitisation and Monetisation Council (DDMC), be renamed AIDAI as it was earlier mandated to study the use and impact of upcoming technologies such as AI on data ethics.

TRAI’s recommendations do not provide an analysis of protections already available in the existing laws, Aggarwal said. They do not analyse the regulatory reforms already in progress, namely the proposed Digital Personal Data Protection Bill (DPDPB) and the ongoing discussions on the Digital India Bill (DIB), he remarked.

ET reported last week that the government has dropped a clause in the revised draft of the DPDPB that enabled AI chatbots and search engines to scrape publicly available personal data. ET also reported that the new DIB, which replaces the 23-year-old IT Act, will have provisions to regulate emerging technologies including AI.

Despite efforts toward regulation by other countries in Europe, particularly Italy, the Indian government has so far maintained its stand to not regulate AI.

ETtech

‘Surfeit of Regulators’

Kazim Rizvi, founder of public policy think-tank The Dialogue, questioned the need for a separate AI regulator for India when the country already has various other regulators in the making, including the proposed “internet regulator” within the upcoming DIB.

Rizvi said that the TRAI-suggested risk-based approach similar to EU AI Act has both positives and negatives.

“As we progress, India must analyse some of the critical aspects of AI and consider drafting enabling regulations to support homegrown AI innovations to serve worldwide,” Rizvi suggested.

The DPDP Bill 2023 which is going to be introduced in the Parliament looks into data digitalisation from a privacy protection angle.

“The recommendation that TRAI be made the regulator on AI-related matters and be given a legal framework, appears to be premature,” Aggarwal said.

However, Nasscom welcomed the recommendations on how to enhance adoption of AI in the telecom sector, expand AI-specific infrastructure and setting up of experimental campuses.

“We also welcome TRAI’s recommendations on an inter-ministerial and inter-departmental collaborative approach to develop skilling both at the academic and professional level,” Aggarwal said.

Building the ecosystem

TRAI’s recommendations to regulate AI cover a wide range of topics ranging from developing the ecosystem, skilling and regulations.

Attention must be given towards enhancing the state capacity and skills in enforcement for effective governance of evolving and niche technologies like AI, Rizvi opined.

At the functional level, the recommendations emphasise the importance of inter-regulatory coordination at different levels, i.e., federal, regulatory, and multilateral levels.

“However, mixing AI and data governance might overlap and make the scope of the regulation too broad,” Rizvi pointed out.

The government is already working on the National Data Governance Framework, which talks of unlocking the value of data that the government holds.

Therefore, the scope of any future regulatory framework of AI must be narrow while various other correlated aspects like data governance, online safety etc., handled under different legislations must be coordinated, he said.

At the operational level, the suggested recommendations are a step in the right direction as they, in a way, tackle the state capacity issue by suggesting alternative market mechanisms like certification, ethical codes etc., to ensure the players follow principles appropriately, he said.

TRAI has also called on the government to collaborate with international agencies and governments of other countries to form a global agency that acts as the primary international body for development, standardisation and responsible use of AI.