AI as the new OS: Can enterprises trust a thinking machine?

Time of India

5 hours ago

Artificial Intelligence has long been touted as a transformative force, but the discussion at the ETCIO Annual Conclave 2025 pushed that narrative into more provocative territory: Is AI becoming the cognitive operating system of the enterprise—and can we trust it to think?
Moderated by Shipra Malhotra, the session brought together Navendu Agarwal, Group CIO, Ola; Shuchi Mahajan, SVP & Head - Fraud Prevention, Analytics, Digital & Customer Awareness, HDFC Bank; Binit Jha, CDO, IDBI Bank, for a candid conversation on what happens when AI becomes not just an assistant, but the actual decision-making engine of the enterprise. From explainability and regulatory friction to cultural shifts and cognitive accountability, the panel explored the real-world implications of enterprise-wide AI adoption.
From dashboards to decisions: A paradigm shift
Navendu Agarwal offered a compelling metaphor to frame the shift: 'AI is no longer a tool—it's the second brain of the enterprise.' Previously, dashboards and BI tools supported decision-making. Today, AI does the heavy lifting of assimilating data across systems and summarizing it contextually, empowering decision-makers directly. This shift redefines agility: enterprise leaders no longer have to depend on layers of interpretation—they can interface with AI directly to derive business value.
Agarwal argued that generative models have fundamentally altered the nature of enterprise intelligence. LLMs can now analyze and interpret massive data volumes without needing task-specific training. 'It takes the power away from technologists and gives it back to decision-makers,' he said.
Regulation vs. innovation: BFSI's balancing act
In the highly regulated world of banking, Binit Jha described AI's rapid evolution as both a boon and a bottleneck. 'AI moves every 24 hours. Banks and regulators don't,' he quipped. For institutions like IDBI Bank—both regulated and a public-sector entity—the challenge lies in matching AI's pace within the constraints of compliance.
Yet Jha emphasized that AI isn't optional. The bank is categorizing use cases based on risk and regulatory clarity to accelerate safe implementation. 'We have to implement it anyway,' he said. The key is to identify areas where AI can operate within existing guardrails—like analytics, scoring, and workflow automation—while human oversight governs final decisions.
Trust is the new OS: Why AI must be explainable
A 2024 PwC India study cited during the session revealed a trust gap: 73% of Indian enterprises trust AI for operational decisions, but only 28% trust it for strategic or financial ones. The difference? Explainability.
Shuchi Mahajan emphasized that while AI augments speed and efficiency—especially in fraud analytics and customer scoring—it cannot replace human judgment where empathy, ethics, and reputational risk are involved. 'Cognitive accountability will become the soul of this new OS,' she said. The future belongs to organizations that embed trust as a cultural cornerstone.
She advocated for a 360-degree trust model that begins with employees. As AI introduces uncertainty about job security, Mahajan stressed the need for leaders to upskill teams not in coding, but in applied AI fluency. 'People must know how to prompt and use AI to solve problems—not build it,' she noted. That shift requires cultural transformation as much as technical modernization.
Defining the red lines: Where AI must not decide
Both Mahajan and Jha outlined clear boundaries where AI should never operate autonomously:
Safety: Anything involving physical or human safety must have human oversight.Reputation: AI decisions impacting organizational credibility need a conscience layer.Bias: Legacy data and opaque models can entrench hidden biases.
Navendu Agarwal added the nuance: 'AI should never replace cognitive work—it should accelerate it.' He underscored that decision-making must stay human unless it's entirely predictable and rule-based. 'AI is your intern—it delivers. But interpretation is your job.'
Redefining leadership for the AI-driven enterprise
Looking five years ahead, the panel explored what strategic shifts CXOs must undergo. Mahajan asserted that tech-first leadership must evolve into trust-building leadership—across employees, partners, regulators, and customers.
Jha suggested a three-stage roadmap:
Self-Assessment: Understand your current capabilities, infrastructure, and team readiness.Tactical Rollouts: Focus on safe, efficiency-enhancing use cases with high trust value.Enterprise-Scale Vision: Build a 10-year view of AI-driven transformation, emphasizing ROI, flexibility, and speed.
Navendu Agarwal agreed but reframed AI as an accelerator, not a replacement. 'Technology becomes more accessible, less intimidating. You don't need a tech PhD to lead transformation—you need curiosity and clarity,' he said. This will democratize innovation, enable faster decision-making, and redefine the CXO role as strategic orchestrator rather than operational executor.
The session concluded with a powerful truth: AI may think, but humans remain responsible. As Mahajan put it, 'If AI is the OS, accountability is the conscience.' Success will belong to enterprises that view AI not as a plug-and-play solution, but as a deeply integrated partner—governed by values, ethics, and strategic clarity.
AI is no longer just a tool—it's becoming the enterprise's cognitive core. But trust, governance, and strategic leadership are non-negotiable. Enterprises must move from AI adoption to AI accountability—balancing speed with ethics, and intelligence with empathy.