Technolegal approach of India on AI pragmatic: Wipro's Ivana Bartoletti

India's techno-legal AI governance model, built on soft guidelines and shared responsibility, is practical for rapidly evolving technology, says Wipro's Global Chief Privacy and AI Governance Officer

India’s techno-legal approach of having soft regulations and suggestive frameworks for artificial intelligence (AI) governance is pragmatic, as the technology will change its nature every few years over the next decade, said Ivana Bartoletti, Wipro’s Global Chief Privacy and AI Governance Officer.

 

“The techno-legal approach adopted by India means that instead of having a purely legal approach, the country looks at accountability across the whole AI lifecycle, issuing soft guidelines and legislation on responsible AI and security requirements,” Bartoletti said. She was in Delhi recently to attend the AI Impact Summit.

 

This approach is more practical because no one can predict where AI will eventually end up over the next few years. Companies in India should approach AI governance through the lens of building it up as a component from the ground up, rather than treating it as a compliance checkpoint or simply a risk-management feature, she said.

 

“Companies need to think of AI governance as a strategic capability if they want to have long-term sustainable innovation. Risk management is only one of the aspects of governance,” Bartoletti said.

 

However, governance responsibilities should be flexible, with both companies and the government sharing responsibilities as needed, she added.

 

“For example, you cannot place all the responsibility on parents to monitor what their children do online. It is impossible. It will require constant adjustment and a broader framework where the basic principles remain unchanged,” she said.

 

Though there has been a concentration of AI power in countries such as the US and China, which has allowed the private sector, academia and venture capitalists to collaborate and create large, powerful AI companies, countries of the Global South, including India, need to demand large language models that represent the country's cultural norms, she said.

 

Policy measures, such as offering long-term tax breaks to global companies willing to set up data centres in India, indicate that the country is not averse to big tech expanding its operations there. At the same time, however, India should seek greater representation of its people and culture to achieve true sovereign AI capabilities, Bartoletti said.

 

Apart from LLM capabilities, India also needs AI applications that are relevant, adaptable, and scalable for its users, she said.