Even as artificial intelligence is being deployed for various use cases, regulations are not in place in many geographies including India. As AI technologies rapidly advance, the absence of regulatory frameworks can lead to misuse, privacy breaches, and ethical dilemmas. businessline spoke with Eunice Huang, Head of AI and Emerging Tech Policy, Google APAC to understand how to prevent a dystopian future through a collaborative approach to establish international agreements and standards for AI development and use
There is a lot of discussion around AI regulations. On one hand there are industry alliances being formed in a bid to self-regulate and on the other hand governments worldwide are looking to implement regulations to limit possible misuse of AI. How do you see regulations evolving going forward ?
Google was one of the first companies to say that AI is too important not to be regulated. However, this has to be balanced with the need to promote innovation. We do recommend that governments look at the existing body of laws to see what already applies and the incremental risks that AI brings. Then, you target that gap so that you don’t have a whole body of laws that make it difficult for companies to navigate and add a lot of friction to Innovation. We also promote international alignment with standards because ultimately it needs to flow between countries. Having a regime that is in line with international standards and international best practices would greatly help, especially Indian start-ups. They are looking not just to serve in their Indian market but also to go abroad.
Some players have called for a pause in deploying AI until necessary guardrails are put in place. Is this feasible now that AI has been rolled out so extensively?
We can walk and chew gum at the same time. There are so many positive and beneficial applications of AI and we shouldn’t throw the baby out of the bath. I think we can find a compromise where we put in the necessary guardrails at the same time allowing the innovation to continue to flow. I think a lot of the onus is also on companies to act responsibly to say that we have certain processes in place to make sure that when we develop AI or we deploy it, we do it responsibly.
However, regulations are falling behind AI developments just like how governments failed to put in place guardrails to check misuse of social media. Doesn’t that leave users of AI at risk?
From my vantage point, compared to social media regulation, we’ve seen governments move at great velocity on AI regulations. The whole rush came in end of 2022 when Chat GPT dropped and since then EU has put in laws and the US has issued executive orders on AI. In Asia, there are about eight to ten different countries that are already processing AI regulations. We also believe that a lot of existing laws already ensure consumer protection. The way new laws are written, it’s very difficult to make sure that regulation keeps pace (with tech developments) so we always recommend that when new rules are written, the regulation should be neutral. It should be able to evolve regardless of how technology evolves.
India is also putting in place AI regulations. What would be your recommendations to Indian policymakers ?
We have three key recommendations. The first recommendation is about investing in innovation infrastructure. This includes the hot infrastructure such as the compute infrastructure that you need to train these models, and also the soft infrastructure which means having an enabling policy environment. Do you regulate based on a risk-based approach? Are you regulating the end applications rather than the underlying technology? Are you looking at international standards? And do you have the right framework in terms of copyright law, privacy law, and cross-border data flow transfer laws to enable the development of AI and deployment of AI in India? The second pillar is really about building an AI-ready workforce. Investing in skills so that people can use AI, and they can benefit from AI, from students to workers to public sector officials. There’s going to be displacement in some areas but there will be new job creation. So how do we equip people to have the relevant skills, to upskill and to pivot? And then the last pillar is about promoting inclusive adoption and accessibility.
