Independent body, risk-based framework: TRAI suggests regulations for AI

India should play a leading role in shaping global AI standards and governance structures, said Trai

The Telecom Regulatory Authority of India (Trai) on Thursday called on the government to adopt a holistic regulatory framework for the responsible development of cutting-edge artificial intelligence (AI) technologies, reported Economic Times (ET).

It has proposed the immediate establishment of an independent statutory authority, 'Artificial Intelligence and Data Authority of India (AIDAI),' for the development of responsible AI and the regulation of use cases in the country.

To expedite matters, it has suggested renaming its previously recommended statutory body, Data Digitisation & Monetisation Council (DDMC), AIDAI, as it was previously mandated to study the use and impact of emerging technologies such as AI on data ethics.

“India should play a leading role in shaping global AI standards and governance structures, considering the sensitivity and far-reaching impact of AI across the nations, which defies borders,” Trai said in its recommendations.

The recommendations followed a consultation paper titled "Leveraging Artificial Intelligence and Big Data in the Telecommunications Sector."

Also Read: Need independent statutory authority to develop, regulate AI: TRAI

Trai called on the government to work with international agencies and governments from other countries to establish a global agency that would serve as the primary international body for AI development, standardisation, and responsible use.

Trai also recommended that specific AI use cases be regulated on a risk-based framework, with high-risk use cases directly impacting humans regulated through legally binding obligations.

According to the statement, India has already played a key role in shaping the G20 AI principles and is now poised to take a leadership role in AI through its G20 presidency.

“Trai’s recommendations are noteworthy as telecom regulators have not developed a regulatory framework for AI in other countries where they were either developed by data protection authorities or where sectoral regulators have been empowered to come up with their own AI framework…the likes of OFCOM, FCC, IMDA or OFCA have not intervened in this space,” a senior executive of a Big 3 telco told ET.

However, the software industry body Nasscom opposed Trai's move to regulate AI.

Also Read: TRAI directs telcos to deploy AI system to crack down on spam messages

"At Nasscom, we believe that while the development and deployment of AI will require oversight to ensure responsible use and scale-up of adoption, we strongly recommend against creating more laws or statutory authority to regulate AI," said Ashish Aggarwal, vice president at software industry body Nasscom.

"Instead, we should ensure that the existing regulatory framework and enforcement capacity are updated to take into account the at-scale deployment of AI," he added.

Trai noted that because the formation of too many statutory authorities or bodies tends to create confusion, the work of AIDAI should be entrusted to the telecom regulator with appropriate modifications to the Trai Act.

It has also proposed the formation of a multi-stakeholder body (MSB) to serve as an advisory wing to the proposed AIDAI.

The telecom watchdog believes that the use of AI products should not be limited to geographical boundaries. Any framework for responsible AI, it stated, cannot work in isolation and will necessitate global collaborations.

The regulator also believes that the rapid deployment of 5G networks in India will accelerate the adoption of AI applications.