A group of experts appointed by the Ministry of Electronics and Information Technology (MeitY) has recommended an interagency or whole-of-government approach to regulating artificial intelligence (AI).
This is because a joint and coordinated approach by different regulatory bodies may allow them to be more effective and efficient, given the complexity that can be associated with dealing with AI systems at scale, according to experts.
According to experts, the committee could be headed by a chief scientific advisor and include representatives from all ministries. Even informal members could be individuals capable of representing the interests of AI developers, AI implementers, data providers, data managers, and end users, so that the perspectives of the entire ecosystem can be taken into account, the expert group said in its report.
The recommendations were outlined in the Report on the Development of AI Governance Guidelines. The report has been put out for public discussion, with comments available until January 27.
The expert group wants MeitY to have a technical secretariat or advisory body and act as a focal point for this governance group, which should create a database of AI-related incidents to understand the actual frequency of AI-related risks in India.
“The call for a whole-of-government approach recognizes the sectoral nuances of AI. With use cases arising in sectors such as healthcare, finance, agriculture, education, logistics, etc., a coordinated, specialized strategy is vital to ensure that AI governance is not only comprehensive but also adaptable to diverse sectoral needs,” said Jamila Sahiba, Senior Program Manager at Dialogue.
According to Ms. Sahiba, it will be important to determine during the consultations how to develop a sectoral framework that takes into account the unique challenges faced by each sector while ensuring that the overall governance principles are consistent.
For the time being, MeitY has no plans to regulate AI, but will instead maintain a general code of conduct until the market matures.
To increase transparency and governance in the AI ecosystem, the panel recommended that the technical secretariat engage the industry in voluntary commitments to transparency in the AI ecosystem as a whole.
The technical secretariat should explore the suitability of technological measures to address the risks associated with artificial intelligence, such as the use of watermarking, labeling, and fact-checking to combat “deep fakes.”
