Infosys EVP Breaks Ground on AI Adoption and OpenAI Partnership

Balakrishna (Bali) D.R., EVP and Global Services Head of AI and Industry Verticals at Infosys

Infosys EVP Balakrishna (Bali) D.R. on what separates successful AI adopters from those struggling to scale and the company's collaboration with OpenAI

Enterprise AI adoption is moving beyond experimentation, but many organisations remain caught between promising pilots and the challenge of delivering measurable business value.

Fragmented data, legacy systems, evolving governance and unclear ownership can all prevent AI initiatives from progressing into sustained transformation, making operational discipline as important as access to advanced models.

Here, Balakrishna (Bali) D.R., EVP and Global Services Head of AI and Industry Verticals at Infosys, explores what separates successful adopters from those struggling to scale, and delves intoInfosys's strategic collaboration with OpenAI.

Why are so many organisations still struggling to scale AI beyond pilots?

Many organisations are not struggling because the AI models are weak; they are struggling because the enterprise foundations need to be better. Too often, AI starts as a cluster of promising pilots, but the underlying data is fragmented, legacy systems are hard to integrate, governance is still evolving and ownership of outcomes is unclear. That is what creates pilot purgatory.

Our own research tells the story clearly. The Infosys AI Business Value Radar, which surveyed senior executives across more than 3,200 organisations across the globe, found that only around 20% of AI use cases are achieving most or all of their business objectives, and a fifth of all deployments are generating no value at all or are cancelled after deployment.

The lesson is straightforward: scaling AI requires an operating model, not a lab model. You need business sponsorship, modern data foundations, clear guardrails and use cases tied to revenue growth, cost optimisation or risk reduction. Otherwise, pilots stay interesting but never become significant.

What separates successful AI adopters from those falling behind?

The organisations pulling ahead treat AI as a CEO-level transformation programme rather than a series of isolated experiments. They align leadership around a clear enterprise strategy, invest early in data readiness and modern architecture, and put governance in place before they try to industrialise use cases. Just as important, they redesign workflows instead of simply layering AI on top of broken processes.

Too often, AI starts as a cluster of promising pilots, but the underlying data is fragmented

Balakrishna (Bali) D.R., EVP and Global Services Head of AI and Industry Verticals at Infosys

McKinsey’s 2025 researchis useful here: AI high performers are nearly three times more likely to redesign workflows and three times more likely than their peers to strongly agree that senior leaders at their organisation demonstrate ownership and commitment to their AI initiatives. High-maturity organisations are more likely to keep AI initiatives in production for three years or more, which tells you they are building for sustained value, not short-lived pilots.

In practice, the winners also bring their workforce with them. They invest in change management, AI fluency and human-in-the-loop operating models so that trust grows alongside adoption. The dividing line is no longer access to AI. It is the discipline to operationalise it.

How will the Infosys–OpenAI partnership help organisations turn AI experiments into measurable business results?

What makes the Infosys–OpenAI partnership strategically important is that it is an operating model for unlocking AI value at scale. The collaboration combines OpenAI’s frontier models with Infosys Topaz Fabricto help our clients move from experimentation to practical, responsible deployment and measurable business outcomes.

The initial focus areas include software engineering, legacy modernisation, DevOps automation, e-commerce and other engineering-led domains. That matters because these are areas where enterprises can usually measure value in improved engineering productivity, accelerated delivery and reduced time-to-market.

Equally important, the partnership is being presented with enterprise governance and responsible deployment at the centre, not as an afterthought. For clients, that means combining world-class model capability with Infosys’ domain expertise, delivery scale and modernisation experience. In other words, the ambition is not simply to add AI to workflows, but to redesign workflows so AI can deliver measurable business outcomes.

Key Statistics

• 3,240 – number of organisations surveyed by Infosys's AI Business Value Radar across 132 AI use cases
• 19% – proportion of AI use cases which deliver on all their business objectives (Infosys)
• 60% faster – Infosys used an AI-first approach to modernise nearly three million lines of legacy code for a global car rental company, completing the transformation 60% faster
• 3 x more likely – McKinsey found AI high performers are nearly three times as likely to fundamentally redesign workflows as other organisations

Infosys recently formed a strategic collaboration with OpenAI

How is AI changing software development and modernisation?

AI is changing software development in two ways at once: it is reducing the effort required for routine engineering tasks, while expanding the scope of what engineering teams can modernise and transform. We are seeing clear value in code generation, documentation, testing, debugging, code review and legacy transformation, but the bigger shift is that engineers can spend more time on architecture, integration, resilience and business logic.

At a global leader in car rentals, we used an AI-first approach to accelerate legacy transformation. We analysed nearly three million lines of legacy code with multimodal LLMs to extract core business logic and modernise it into a scalable, cloud-based platform 60% faster. This approach reduced hosting costs, increased the reuse of business capabilities across channels and created a scalable foundation for future modernisation and automation.

Our partnerships with AI natives and hyperscalers all point in the same direction: AI-assisted engineering is moving from isolated coding support to end-to-end software delivery and legacy modernisation. That raises the bar for engineering talent, but it also elevates the role of engineers in creating business outcomes.

What makes the Infosys–OpenAI partnership strategically important is that it is an operating model for unlocking AI value at scale

Balakrishna (Bali) D.R., EVP, Global Services Head, AI and Industry Verticals at Infosys

What will define the next phase of enterprise AI adoption?

The next phase will be defined by orchestration. We are moving beyond chat interfaces and isolated copilots towards AI agents, specialised models and enterprise-wide systems that can reason, coordinate and act across workflows.

But success will not come from autonomy alone. It will come from combining agents with context engineering, governed data access modernised estates and clear human accountability. That is why I believe the winners will be organisations that build trusted, multi-model AI ecosystems rather than betting on a single tool or vendor.

We have been clear on this direction from the start. Topaz Fabric is designed to unify models, data, agents and workflows, while our broader partner ecosystem – spanning from silicon to apps – reflects a deliberately open, composable approach.

Industry evidence points the same way. We are hearing that agentic AI interest is rising, but deep scaling remains limited. The World Economic Forumargues trust and governance are the real constraints for adoption; otherreportsshow AI is becoming affordable and more accessible, which means competitive advantage will increasingly come from execution, not access. The next phase of enterprise AI will, therefore, be less about experimentation and more about redesigning the enterprise itself.

Infosys's key AI partners

• OpenAI:Strategic partner for enterprise AI transformation, including software engineering, legacy modernisation and agentic AI.
• Microsoft:Partner across Azure, Azure OpenAI Service and AI-powered enterprise solutions.
• NVIDIA:Collaboration focused on AI infrastructure, models, platforms and enterprise AI deployment.
• Google Cloud:Partner supporting generative AI solutions through Google Cloud technologies and Infosys Topaz.
• AWS:Collaboration combining Infosys Topaz with AWS AI capabilities to accelerate enterprise adoption.