AI governance, checks and balances critical to building trust in autonomous agents: Axis Max Life executive
New Delhi, August 7
Trust, governance and clearly defined safeguards will become increasingly important as businesses move towards agentic artificial intelligence, where autonomous AI agents can make decisions and take actions, according to Sanjay Thawakar, Head - AI, Data, Innovation and BIIG at Axis Max Life Insurance.
Speaking to ANI on the sidelines of the 3rd India International Fintech Festival, Thawakar said discussions around AI are increasingly moving towards agentic AI and agentic commerce, raising questions about how consumers can trust autonomous systems with decisions and actions.
"When we think about agents. How do we build this trust? How do we build that people do believe that most of these agents are going to work in the manner that they have to, and they can trust them with making decisions and acting on it," he said.
Thawakar said governance will play an important role in building this trust, particularly as AI systems become more autonomous.
"What are the kind of authorisations that it can have? What is the kind of transparency in terms of decision-making? And what are the areas that it can go and do those transactions in?" he said.
"The important thing that comes out is that the entire governance platform - how do we govern the AI? How do we ensure that there are certain checks and balances in place? So it's a very important leg in this, in building that trust," Thawakar added.
He said cybersecurity should be viewed as part of a broader governance and risk-management framework, especially as autonomous agents begin interacting with the open web.
"When we are deploying these particular agents, and they will start having an action in terms of interacting with the open web, then the question comes down to what are the kind of checks that we are putting in? What are the kind of cyber security principles that we are putting in?" he said.
"It's a bigger umbrella of risk and governance. Cyber security just becomes a part of it," he added.
On the impact of AI on employees, Thawakar said organisations also need to focus on human-AI coexistence and change management as automation expands.
"The core of it is that a coexistence with humans," he said, adding that organisations need to train employees and help them understand how the technology can take over mundane work.
"The mundane part will be taken over, so it leaves you with enough time to do that exciting thing," Thawakar said.
— ANI
Reader Comments
The point about cybersecurity being part of a bigger risk governance umbrella is spot on. We can't treat cyber threats as standalone challenges anymore when AI agents will be roaming freely online. India needs a robust framework before we start letting these autonomous systems loose on our banking systems.
As someone working in fintech, I appreciate this perspective. But honestly, the bigger challenge in India is digital literacy. If people can't fully trust a bank app, how will they trust an autonomous AI agent handling their insurance or investments? We need to build trust at the basic level first. Still, good direction to be thinking about. 👍
This sounds good in theory, but I'm skeptical about how much governance will actually be enforced. We've seen in the US that tech companies love talking about ethics frameworks while scaling unregulated AI. India has a chance to do this right from the start, but only if it's more than empty promises.
The human-AI coexistence aspect really hit home for me. I've seen so many colleagues worried about automation taking their jobs, especially in the insurance sector where data processing is huge. But Mr. Thawakar is right — if we use AI to handle the boring repetitive work, we can focus on actually helping customers with meaningful things. The mindset shift is crucial.
Interesting to see an insurance company talking about AI governance. But here's my concern: when these AI agents start making decisions about claims, who's accountable? The customer breaking a mobile on the road wants to know if an AI or a human decided to reject their claim. Governance frameworks need to define clear accountability chains, or this trust-building exercise will
We welcome thoughtful discussions from our readers. Please keep comments respectful and on-topic.