AI's next chapter will be defined by trust, governance and outcomes, not bigger models

New Delhi, July 16 -- As enterprises move beyond AI experimentation, the debate is no longer whether businesses should adopt artificial intelligence. Instead, the focus has shifted to how AI should be governed, where it delivers the greatest business value, and how organisations can deploy it without compromising trust. That's one message emerging from enterprise technology leaders this AI Appreciation Day, observed annually on 16 July.

Across industries, executives say the next phase of AI will be less about building larger foundation models and more about creating secure, domain-aware and human-centric AI systems that enterprises can confidently embed into their core operations.

From AI adoption to AI execution

"The conversation has entirely shifted to execution," said Dan Mountstephen, General Manager and Senior Vice President, Asia Pacific and Japan at Okta. "No enterprise is going to rely on a single AI platform, model or cloud vendor. The real challenge is consistent governance."

As organisations deploy multiple AI models across different business functions, Mountstephen argues that identity management is becoming the control plane for enterprise AI. Human users are no longer the only entities requiring governance, AI agents, APIs and service accounts now need the same visibility, lifecycle management and access controls to ensure they operate safely.

That focus on governance was echoed across several technology leaders, who argued that responsible AI is becoming a business necessity rather than a compliance exercise. "Responsible AI must become an integral part of the engineering process rather than an afterthought," said Chandan Govindarajulu, Executive Vice President, Global Delivery & Business Leader at Virtusa. Building trustworthy AI, he said, requires transparency, rigorous testing and continuous human oversight so organisations can confidently deploy AI at scale.

For many executives, the next competitive advantage will not come from faster adoption alone but from combining AI with human judgement.

"AI is no longer just recording what happened. It understands context, anticipates what comes next and takes action," said Saurabh Saxena, Intuit India Site Leader and Senior Vice President, Data, Growth & Experiences. However, he cautioned that autonomous AI systems will only succeed when supported by trusted data, deep domain expertise, transparency and human oversight.

A similar view came from Tushar Agnihotri, CEO of Route Mobile, who said AI has moved beyond being purely a technology discussion. "Artificial intelligence is no longer a technology conversation; it has become a boardroom imperative," Agnihotri said. Organisations that succeed, he added, will be those that embed intelligence into everyday customer interactions and business decisions while maintaining governance, transparency and trust.

Several leaders also believe enterprises are entering the era of agentic AI, where AI systems evolve beyond assisting employees to actively collaborating in workflows and business processes. According to Venugopal Ganganna, Co-Founder and CIO of LS Digital, organisations are moving from using AI as a productivity tool to deploying it as an active collaborator capable of orchestrating customer journeys, optimising investments and supporting faster decision-making.

"The organisations that succeed will be those that design workflows where people and AI work together to create better outcomes," he said, adding that AI-native businesses will require strong data foundations, connected technology platforms and responsible governance.

Trust, governance and security take centre stage

Security is becoming equally central to that discussion as AI systems gain greater autonomy. "The future of AI will not be decided by who builds the biggest model," said Subir Sangal, Chief Executive Officer of Eagle Information Systems. "It will be decided by who builds AI that understands the business it is meant to serve, and keeps it secure."

He noted that cybercriminals are increasingly using AI to automate phishing, malware development and cyberattacks, making enterprise security and continuous monitoring indispensable.

That concern is shared by cybersecurity vendors, who argue that AI adoption cannot be separated from risk management.

"Responsible, transparent and well-governed AI is not just an innovation imperative, but a national cybersecurity imperative," said Dr. Sanjay Katkar, Joint Managing Director at Quick Heal Technologies. With AI-assisted phishing and identity attacks accelerating, he said organisations must ensure AI strengthens trust rather than creating new vulnerabilities.

Infra and private AI - the next frontier

While governance dominated the discussion, infrastructure also emerged as a defining factor for AI's future.

"The future of AI will be built on resilient, scalable and sustainable networks," said Badri Gomatam, Group Chief Technology Officer at STL, arguing that enterprise AI ambitions will ultimately depend on the quality of the digital infrastructure supporting them.

Others believe enterprises must rethink where AI actually runs. Praveer Kochhar, Co-Founder and Chief Product Officer of KOGO AI, argued that businesses are increasingly favouring private AI deployments that keep models, enterprise data and institutional knowledge within their own infrastructure rather than relying exclusively on public AI platforms. Trust, he said, depends on maintaining control over data movement, model selection and auditability.

The emphasis on trust extends beyond large enterprises. Brijesh Agarwal, CEO of Busy Infotech, said India's small businesses will adopt AI only if it remains explainable, secure and affordable. "AI should support human judgement, not replace it," Agarwal said, adding that its value should ultimately be measured by time saved, errors prevented and better decisions.

Taken together, the industry's message reflects a broader shift in enterprise AI priorities. The excitement surrounding large language models is giving way to more practical questions around governance, security, infrastructure, domain expertise and measurable outcomes. That said, AI may still be evolving rapidly, but for enterprises, the defining challenge is no longer building smarter machines. It is building AI that organisations, employees and customers can trust.

Published by HT Digital Content Services with permission from TechCircle.