Generative AI adoption is one of the most actively debated topics across corporate circles today. While many independent studies indicate that organisations are planning to significantly increase AI spending, Gartner says there is room for more and if India does not act fast, we might miss the bus. According to them the regulatory framework is going to be a key driver in this context.
Need for more AI Security startups
Speaking to Financial Express Online on the sidelines of the Gartner Security & Risk Management Summit held in Mumbai recently, Manjunath Bhat, VP Analyst, Gartner pointed out that he “would have loved to see more innovation coming from cybersecurity providers in India who can mitigate generative AI security risks. Unfortunately, if you think about it, there are very few startups that are focused on AI security. A lot of it is still in the good old API security or application security, but they have not evolved to offer Gen generative AI security.”
This according to him is quite contradictory to the situation elsewhere and you now have “vendors in Europe, Israel and North America. So, I feel like that’s something for India to see and maybe even tap into this opportunity. Startups should not miss out on this. If we don’t invest in it, we’ll also see funding sort of dry up. Globally, all of the funding is going to cybersecurity startups and more so AI security startups. And if India is not India is not considered a player. We miss out on that.”
Regulatory framework crucial
However, in this context Bhat believes that regulatory frameworks are going to be the driver for securing Gen AI. “I think regulatory frameworks are playing a huge role. Because if you think about it left to their own devices, it’s very difficult and in many cases, even for the providers to understand how do you mitigate this? Where do you draw the line? What’s considered acceptable and what’s considered not acceptable?,” he added, outlining the importance of well thought out policies that do not come in the way of investment.
Giving examples of some classic risks like face recognition Bhat explained how there is a dilemma where “face recognition be allowed in certain contexts and where it is not allowed in some other modelling site. And it’s not for the providers to decide. Because it might lead to conflict of interest, depending on which businesses they serve, and also where the public sector is using. So regulatory frameworks, I think, in the context of Gen AI, are going to be the driver for securing Gen AI.”
He takes a step further and expects that by 2030, “the biggest influence from a security perspective will be in terms of the regulatory framework. So, I think that’s so far, in 2024, these regulatory frameworks are just emerging. So, by 2030, you’d have organisations who have already embraced these frameworks and implemented them.”
Roadmap for 2030
So in terms of a roadmap for 2030, what’s the outlook exactly? According to Bhat, “By 2030, you can expect quite a few things. One is from a security perspective, auto remediation would have become unknown.” He expects auto remediation of security vulnerabilities to experience some kind of “self healing” over the next few years.
On the development side, he is looking forward to “applications being developed just based on prompts. So that’s quite feasible. Another thing by 2030, would be you will really start to see people who become conversant in multiple programming languages. Gen AI enables cross-skilling. So, by 2030 people become much more conversant in a multitude of programming languages.”
