At the heart of the discussion was a more expansive understanding of ROI. Return on investment remains important, but enterprises are increasingly looking for something broader. They are looking for return on intelligence. This means better decisions, stronger resilience, new business models, improved customer engagement and work that is redesigned for a future where humans and AI systems operate together.
Setting the context for Return on IntelligenceRomal Shetty, CEO, Deloitte South Asia, opened the Forum by framing AI ROI as a leadership and enterprise value question. The opening segment moved the discussion beyond the narrow view of AI as another technology investment or cost optimisation programme.
The larger question was whether AI is only helping organisations save costs or whether it is amplifying intelligence across the enterprise. That distinction shaped the tone for the Forum. AI value cannot be limited to efficiency alone. It must also be assessed by the quality of decisions it enables, the resilience it builds and the new possibilities it opens for growth.
For India, this perspective has particular significance. Scale offers a major advantage, but scale by itself does not create enterprise value. The discussion placed trust and governance, human led AI powered transformation and return on intelligence at the centre of India’s opportunity. The message was clear. AI can change how corporations work and how the country builds capability, but only when ambition is matched with responsible execution.
The global value pools of AI
The next stage of the Forum moved from ambition to execution. Jim Rowan, Head of AI, Deloitte US, led the discussion on what it will take for organisations to scale AI adoption and connect investment with measurable business value.
The session framed AI value across three broad areas. Efficiency AI continues to receive significant enterprise attention as organisations look to automate work, improve productivity and reduce friction in operations. Experience AI is reshaping how companies engage with customers and employees through chatbots, avatars, video rendering and new forms of AI enabled interaction.
The third area, Growth AI, carried a more transformational promise. It focused on how organisations can use AI to reimagine business models, enter new markets and open new routes to growth. While efficiency remains the most immediate area of investment, the Forum positioned growth as the next frontier for enterprises that want AI to deliver strategic advantage rather than incremental improvement.
This part of the Forum also reflected the pace of change. With frontier model companies, hyperscalers and technology providers releasing new capabilities at speed, enterprise leaders are being forced to separate hype from reality. The challenge is not only to invest in AI, but to know which investments matter, how they connect to business value and how that value can be explained to boards and executive leadership.
Trust, governance and the boardroom shift
As AI scales, the most important decisions are no longer confined to technology teams. Beena Ammanath, Global AI Institute Leader, Deloitte US, led a conversation that placed AI investments, value creation, trust and governance in the boardroom context.
The session explored how enterprises and investors are evaluating AI through the lens of economic value. Cost efficiency remains a major stream of AI adoption, with contact centres, software development and enterprise productivity already seeing clear momentum. But the more compelling question is whether AI can create net new economically valuable output.
This broader lens changes the way organisations assess AI use cases. In sectors such as legal services, for instance, the value may not only come from reducing the time required to review a contract. It may also come from the downstream impact of faster decisions, quicker intellectual property development and improved business velocity.
The governance dimension was equally important. As enterprises experiment with AI, sandbox environments are becoming important spaces for testing, validating and refining use cases. The discussion also highlighted the need to bring compliance, risk and legal teams into the process early, during the pilot phase itself. That shift allows organisations to conduct due diligence before AI moves into production, making trust a foundation for scale rather than a barrier to it.
Measuring AI value across the enterprise
The Forum then turned to the practical challenge of value creation at scale. Prashanth Kaddi, Partner, Deloitte India, led a session that examined how enterprises are using AI implementation across industries and how they can begin measuring returns more effectively.
The discussion showed that AI is now touching almost every part of the enterprise. It is influencing how organisations work, how machines are built, how products and services reach customers and how dealer and partner networks create value. AI agents are being used both in software development and in business workflows. Conversational assistants are helping customers get answers faster, while enterprises look for ways to convert those interactions into leads, orders and stronger customer outcomes.
A more mature ROI framework emerged through four lenses. Direct value captures revenue growth and cost savings. Indirect value looks at benefits that appear upstream or downstream in the value chain. Efficiency value reflects productivity gains and the ability to redeploy people to higher impact work. Opportunity value captures the work that becomes possible because AI changes the economics of time, effort and resources.
This perspective is important because AI value is rarely isolated. A single improvement in one part of the business can create effects elsewhere. As AI measurement matures, organisations will need to move beyond simple calculations of hours saved and begin connecting AI to revenue, resilience, customer conversion, operational efficiency and new opportunity creation.
Scaling agentic AI with discipline
The shift from pilots to scale formed the next major theme. Ashvin Vellody, Partner, Deloitte India, led a session on the strategic choices organisations must make as they scale AI and agentic AI.
Scaling AI is where ambition meets operational reality. Enterprises need to decide what talent to build internally, which hyperscalers and technology partners to work with, what kinds of agents to create and how quickly to respond to obsolescence.
The conversation framed platform decision making around four priorities. Control is critical as agents spread across functions and require governance layers that can monitor, orchestrate and trigger them in structured ways. Context matters because agents need clean, connected and meaningful enterprise data. Connectivity is essential because large organisations operate across hundreds of internal and external systems. Compliance remains central as security, accountability, privacy and explainability become non negotiable in the agentic world.
The session also reflected the changing build versus buy debate, with democratised AI stacks enabling internal capability building while specialised providers continue to move quickly. As enterprises scale, roles and metrics may also evolve, with greater focus on outcomes, customer satisfaction and value delivered, alongside emerging roles such as agent supervisors and AI product managers.
Industry impact and the engineering of AI value
The Forum then widened the lens to industry realities. Vinay Prabhakar, Partner and Leader, Sales, Alliances and Pursuit Excellence, Deloitte South Asia, led a session on how organisations are creating real impact with AI.
A key theme was the need to focus not only on the science of AI, but also on the engineering of AI value. Models matter, whether large, small or domain specific. But the real challenge lies in building effective applications that use those models to solve business problems.
The discussion placed AI in a broader historical context. If the steam engine amplified physical effort, AI has the potential to amplify human intelligence. But as with earlier transformations, its full impact will take time to unfold.
The session also reflected the organisational tension many companies face. AI can generate enthusiasm, pressure from business teams and a surge of pilots across markets and functions. Without discipline, this can lead to duplicated use cases, higher costs, privacy concerns and unclear ROI.
The pathway to value requires structure. Organisations need a clear vision for AI adoption, supported by specific missions, use cases and forums that can evaluate opportunities continuously. The goal is to channel curiosity into initiatives with merit, relevance and measurable value.
The next wave of agentic and physical AI
The next phase of ROI was framed around AI that does not only generate, but acts. Sanghamitra Pati, Applied AI Strategic Growth Market Leader, Deloitte USI, led a session on the rise of agentic AI, physical AI and the infrastructure needed for enterprise transformation.
The discussion reflected a clear shift. It is no longer about if AI will shape the enterprise, but when and how. Three forces are emerging. Agentic and generative AI are transforming industries. Physical AI is opening the path for autonomy in anything that moves. Infrastructure is becoming the foundation that allows these technologies to scale.
Agentic AI was positioned as a defining feature of the future enterprise, where specialised agents may interact with one another much like departments and functions do today. These agents will require humans to work alongside them in new ways.
Physical AI adds another layer as autonomous systems expand beyond cars and robotics. This will demand significant compute power to train models, simulate real world behaviour and deploy intelligent systems reliably. Without powerful, controlled and disciplined infrastructure, enterprise AI ambition cannot translate into scalable transformation.
Redesigning work for AI value
The final segment brought the discussion back to people, work and operating models. Dheeraj Sharma, HR Transformation Leader, Deloitte USI, led the closing conversation on why AI value realisation depends on redesigning work and embedding intelligence into workflows.
The session captured the questions many organisations are asking. Where is the return on technology investment. How can AI be scaled without disrupting what already works. How can enterprises upskill, operationalise AI and become AI first while keeping humans at the centre.
Three common anti patterns shaped the discussion. The first is technology only thinking, where organisations procure tools without reimagining the work those tools are meant to improve. The second is the process trap, where AI is inserted into existing workflows without asking whether the workflow itself should change. The third is a narrow focus on training that is not supported by cultural change.
The more transformative path requires organisations to examine how work actually gets done, including workflows, touchpoints, roles, decisions and collaboration patterns. Since AI is probabilistic and non-deterministic, human judgment, work design and culture become central to value realisation. Models alone will not deliver ROI. Enterprises need operating models that allow humans and AI systems to work together effectively.
From ambition to accountable impact
Deloitte’s AI Forum 2026 made clear that enterprise AI has entered a new phase. The first wave was defined by experimentation and adoption. The next will be defined by outcomes, governance, scale and measurable value.
Across the Forum, the discussion moved from leadership intent to execution, from trust to measurement, from industry use cases to agentic and physical AI, and finally to the redesign of work. Together, these themes showed that AI ROI is not a single metric, but a broader enterprise discipline.
The shift from return on investment to return on intelligence captures the challenge ahead. Intelligence only matters when it delivers impact. Impact only matters when it creates value.
