Key Takeaways
- AI is being regarded as a “second brain” in business, rather than a dashboard tool.
- Banks are fast-tracking AI in areas like scoring and workflow, while keeping final decisions under human control.
- Companies believe in AI for operations but hold back on strategic use due to lack of explainability.
These key insights were further elaborated upon at the ETCIO Annual Conclave 2025, where enterprise leaders collectively acknowledged that AI is rapidly transforming into the cognitive operating system of modern organizations.
Elaborating on this, Navendu Agarwal, Group CIO at Ola, vividly described AI as the ‘second brain of the enterprise’. He explained how AI models can now process and summarise data in ways that once required entire teams. This reduces the lag between data and decision, making businesses more responsive.
Agarwal pointed to large language models (LLMs) as a turning point. These systems don’t need task-specific training, which lowers the barrier for decision-makers outside IT. In his view, this shift is democratising insight and speeding up business response.
BFSI Leaders Call for Safe Innovation
Banking and financial services, bound by heavy regulation, face a complex AI landscape. Binit Jha, Chief Digital Officer at IDBI Bank, summed it up: “AI moves every 24 hours. Banks and regulators don’t.”
Still, he argued that AI adoption is not optional. His team is mapping use cases by risk and compliance levels. AI is being used for scoring and workflow automation, with human oversight on final calls. The goal is safe acceleration – deploy AI where it fits within existing frameworks.
This drive for adoption, particularly in regulated industries like banking, highlights a universal challenge: building trust in AI.
Why Trust Still Holds Back Strategic AI Use?
A recent PwC India survey shows a clear gap. While 73% of businesses trust AI for operations, only 28% trust it for strategic or financial decisions. The difference is explainability.
Shuchi Mahajan, SVP at HDFC Bank, stressed that speed isn’t enough. “Cognitive accountability will become the soul of this new OS,” she said. Especially in fraud analytics and customer scoring, AI must explain its decisions. When it can’t, human judgment must take over.
She also flagged internal trust as vital. As AI enters workflows, employees must feel confident using it. That doesn’t mean learning to code, but learning how to work with AI. Prompting, interpreting, and applying AI will become core skills.
Where AI Must Not Decide
Given the discussions on explainability and human oversight, everyone on the panel agreed there are boundaries AI shouldn’t cross, with human judgment always taking priority.
- Human or physical safety: Always needs human control.
- Reputation and ethics: AI should not make decisions with brand impact.
- Bias risks: Models trained on flawed data can replicate harmful patterns.
Agarwal offered a useful frame: AI is like an intern. It can do the legwork, but the manager must still interpret and decide.
Leadership for the AI Era
The panel outlined what it takes to lead in an AI-driven organisation: clarity, cultural awareness, and a long-term view. Binit Jha offered a practical framework for organisations preparing to scale AI. It begins with a realistic self-assessment of existing capabilities, followed by targeted rollouts of low-risk, high-trust use cases. The final step is to commit to a long-term AI roadmap – one that balances compliance, adaptability, and measurable business outcomes.
Navendu Agarwal also added that the technical complexity of AI is no longer the main barrier. Tools are more accessible than ever. Leaders don’t need to write code—they need to ask the right questions, understand where AI fits, and take responsibility for its outcomes.
The panel closed with a shared view: As Shuchi Mahajan put it, “If AI is the operating system, accountability is the conscience.” The organisations that will thrive are those that treat AI not as a shortcut, but as a partner—one guided by human judgement, ethical boundaries, and long-term thinking. For enterprises at the edge of transformation, the message is clear. Success with AI won’t be defined by speed or scale alone – but by how responsibly and intelligently it is led.
