MeitY's Advisories on AI
A fortnight of regulatory changes
Introduction
On the 1st of March, 2024, India’s Ministry of Electronics and Information Technology (MeitY) sent ripples in the deep tech ecosystem by issuing an advisory note regarding regulation of artificial intelligence (AI) platforms in India. This advisory stated that platforms that use AI models must seek explicit permission from Indian government before deploying it for users in India; in addition, the advisory also stated that such platforms should also label AI generated content as such to avoid being misused by deepfake technologies. However, on 15th of March, 2024, MeitY issued a revised advisory rescinding the requirement to seek permission, while retaining the requirement to label AI generated content. This is a welcome change and the Ministry must be appreciated for taking the feedback of the Industry and Academia and revising the advisory.
Need for the Revision
Even though it was clarified later that the original advisory was targeted at large platforms and does not apply to startups, the advisory spooked many of us in the deep tech ecosystem for the following reasons:
There was no clarity on how the approval process would be. There was no information on who would be responsible for approving, what criteria would they look for and how long the process might take.
Whether it would be possible to objectively evaluate AI models on their ‘appropriateness’? ‘Appropriateness’, is in itself, very subjective and if so, would it be possible to objectively approve or deny permission to AI models?
What if this regulatory change slows down adoption of AI in India? If approval process takes arbitrarily long periods of time, it could extend deployment cycle times for businesses and could disadvantage both businesses and consumers.
While it is extremely important for Government authorities to regulate any technology that has the potential to be misused by people or Organizations, it is equally important to ensure that such regulation is optimal and not over restrictive to slow down innovation. We do not want to be left behind when such tectonic shifts take place in technology and we want both enterprises and consumers to benefit from the advancements in such technology.
Conclusion
The events of this fortnight have emphasised the importance of optimal regulations in an emerging field such as AI. Too much regulation could slow down innovation while too little could lead to unrestricted damage. MeitY showed that its trying to be proactive by making India the first country in the world to bring in regulation on AI, and when the experts felt that the first version of the regulation was suboptimal, the Ministry was quick to revise it showing that it is also flexible and open to feedback. Over the next few years we will see more of the tug of war between the advancement of emerging technologies, people and organizations that want to use it to their benefit at the cost of larger common good and the policy makers who need to frame regulations to restrict such use while encouraging innovation.