Government Clarifies AI Advisory Amid Concerns, Assures Startups Exempted
4th March 2024Union Minister of State for Electronics and Information Technology, Rajeev Chandrasekhar, addressed the recent uproar surrounding the Centre’s directive mandating government approval for the launch of Artificial Intelligence (AI) products. The advisory raised eyebrows, particularly among startups, prompting Chandrasekhar to provide clarity via X.
According to Chandrasekhar, the advisory primarily targets large and significant platforms, excluding startups from its purview. He emphasized three crucial points to dispel confusion surrounding the second advisory under the new IT rules. The directive aims to regulate untested AI platforms from deploying on the Indian internet, ensuring consumer safety through a process of permission-seeking, labeling, and consent-based disclosures.
Highlighting the shared goal of ensuring the safety and trust of India’s internet, Chandrasekhar underscored the government’s commitment to fostering a secure online environment for users and platforms alike.
The government issued the recent advisory on March 1, building upon the initial directive issued in December last year to address deepfake issues on social media platforms. Chandrasekhar cited the need for caution following the Google Gemini AI tool incident, which violated legal norms by generating unlawful content.
Platforms are urged to exercise discipline in transitioning AI models and platforms from labs to the market, ensuring compliance with regulations. Concerned platforms are required to furnish a status report and specify actions taken within 15 days of the advisory.
In response to Google Gemini AI tool’s alleged bias in response generation, Chandrasekhar criticized the tech giant, emphasizing that labeling it as “unreliable” doesn’t absolve it from legal obligations.
The government’s initiative to develop an AI regulation framework, slated for release in June or July 2024, aims to harness AI for economic growth while addressing associated risks and harms.
