AI Models Require Government Approval Before Use

Governments Mandate Approval Process for AI Models: New Regulatory Era Begins

The Ministry of Electronics and Information Technology (MeitY) has issued an advisory that any artificial intelligence (AI) model including generative AI models, which are being tested or deemed to be generally unreliable should be examined for permission from the government before it is exported to the country, media have reported.

The advisory follows recent incidents where users found inaccurate and misleading information in responses to Google’s Gemini AI chatbot about Prime Minister Narendra Modi. As reported by The Economic Times, the advisory was issued on March 1, directing companies to follow its guidelines in detail.

The consultant urged companies that have already deployed AI platforms in the country to ensure that their computing resources do not introduce bias, and discrimination and jeopardize the integrity of the electoral process. Additionally, MeitY is said to have asked AI platforms to amend metadata if AI-generated content could spread misinformation or cause serious falsification.

Union Minister Rajiv Chandrashekhar on Monday gave an insight into the revised advisory on Artificial Intelligence (AI) platforms issued by the Ministry of Electronics and Information Technology (MeitY). He clarified that the advisor is only associated with major platforms and does not include start-ups in the country, as per business norms.

Referring to social media platform X, formerly known as Twitter, the minister stressed that "The recent MeitY advisory should be interpreted accurately. It specifically targets prominent platforms, and the requirement for permission from the ministry applies solely to large platforms, exempting startups."

In addition, companies were asked to include transparency if their platform tends to engage in unreliable behavior and can lead to inaccurate information. Additionally, platforms will need to warn users against using AI to create deepfakes or any information that could influence the election in any way, the report said. While the advice is not currently legally binding, it points towards the direction of AI law in India, going forward.

The Union Minister also clarified that the advisory was aimed at “important events” and only “large events” should seek permission from the MeitY. This advice does not apply to startups. He added that following the consultant’s guidance is in the industry’s best interest because it provides insurance for users who might otherwise file a lawsuit against the platform. “Safety & Trust of India's Internet is a common and unifying goal for governments, users, and forums,” he said.

Conclusion: As governments around the world embark on this journey to AI policy, the ultimate goal is clear: harness the transformative power of AI, ensure that its benefits are distributed equitably and risks are managed role in responsibility by establishing a strong regulatory framework and fostering a culture of responsible innovation establishes an environment in which AI technology serves the interests of all and supports the fundamental principles of fairness, transparency of transparency and accountability.