Update on guidelines for AI developers

On March 5, 2024, Ashwini Vaishnaw, the Indian Minister of Electronics and Information Technology defended the government’s decision to issue guidelines for AI developers. It included tech giants like Google, Microsoft, and Adobe. He emphasized the importance of responsible AI development, stating that platforms must undergo rigorous testing and training to safeguard citizen safety and uphold democratic values. This statement was in response to criticism from both Indian and international tech leaders who argued that such advisories might hinder innovation within the nation.

Government’s directive is advisory

Vaishnaw clarified that the government’s directive is advisory in nature, aimed at encouraging thorough testing of AI models prior to their deployment. “It is imperative for companies to conduct extensive testing and training of their platforms before launch. This will ensure they are safe and reliable for public use. This approach is crucial for the protection of our citizens and the preservation of democratic processes” Vaishnaw asserted.

The issuance of this advisory followed remarks by Rajeev Chandrasekhar, Minister of State for Electronics and Information Technology. Who highlighted a breach of the Intermediary Rules (IT Rules) 2021 by Google’s AI chatbot, Gemini, labeling it as non-compliant due to inappropriate responses.

The Mandate to developers

The advisory mandates the guidelines for AI developers that developers of AI technologies including industry leaders like Google, Microsoft, and Adobe must seek explicit approval from the Indian government for any AI models that are not fully tested or reliable before they are made available to users in India. It also applies to entities that offer platforms powered by unverified or incomplete AI technologies. Those looking to establish a “sandbox” testing environment are required to obtain governmental consent and must clearly indicate that their platform is in a testing phase.

The Concern

The Ministry of Electronics and Information Technology (MeitY) also expressed concerns about the potential for AI-driven biases to undermine the integrity of India’s electoral process. It has asked all affected platforms to report back within 15 days with a detailed account of their compliance efforts and current operational status.

In a subsequent clarification, Chandrasekhar noted that the advisory does not extend to start-ups. This potentially easing concerns among emerging businesses in the AI sector.

In related developments, Smart Group has been proactive in fostering innovation by incubating new AI startups, aiming to contribute to the vibrant ecosystem of technological advancement and ensuring that new ventures have the support they need to navigate the regulatory landscape and thrive in a competitive market.