'AI readiness, not adoption, will define India's enterprise AI success'

After two years being dominated by experimentation, proofs-of-concept and pilot projects, enterprises are discovering that deploying artificial intelligence is significantly easier than turning it into measurable business outcomes. While organizations across sectors are racing to launch AI initiatives, industry leaders warn that many remain unprepared for the operational, governance and organizational changes required to scale those efforts. During a recent roundtable hosted by Avaali Solutions in Bengaluru, enterprise leaders, policymakers and data governance experts debated what they described as India's growing AI execution gap. The discussion comes at a time when artificial intelligence is projected to contribute as much as $1.7 trillion to India's economy by 2035 and government-backed initiatives such as the IndiaAI Mission are accelerating adoption across industries. Yet despite growing investments and widespread enthusiasm, the speakers argued that many organizations continue to struggle with how to convert AI experimentation into business value. Identifying the problem According to Srividya Kannan, Founder and CEO of Avaali Solutions, enterprises are increasingly approaching AI with excitement but often without a clear understanding of where it can create the greatest impact. She believes the first challenge organizations need to solve is not selecting a technology platform or identifying a model, but determining which business problems deserve attention first. Too many companies, she said, continue to rely on brainstorming workshops and isolated innovation exercises rather than objective assessments of where AI investments can generate measurable returns. "AI adoption is no longer the difficult conversation in most boardrooms. The harder question is readiness," Kannan said. "Do enterprises know which processes are worth transforming first, do they have the data and governance to support AI, and can they convert experimentation into measurable outcomes?" Kannan argued that the enterprise prioritization problem has become one of the biggest obstacles to meaningful AI transformation in India. While organizations have launched numerous pilots over the last two years, many have failed to move beyond proof-of-concept stages because they were not linked to clearly defined business objectives, she added. The challenge, according to Sanjeev Kumar Gupta, CEO of the Karnataka Digital Economy Mission (KDEM), is not a lack of infrastructure or talent. India has already built significant AI foundations through investments in compute infrastructure, public datasets and talent development. However, translating those advantages into business outcomes remains difficult. Gupta noted that while a large majority of organizations globally are already using AI in at least one business function, only a small percentage have successfully translated those initiatives into meaningful business value. "To bridge the gap and achieve real business value, enterprises must overcome three core operational barriers: real data availability, hybrid skills that blend domain expertise with AI, and bridging the boardroom understanding gap so directors can actively steer investments," Gupta said. Critical problems Instead of asking where AI can be applied, enterprises should first determine which business problems are most critical, what impact solving them could create and how success will be measured, she said. In her view, the industry has reached a point where continuing to launch large volumes of pilots offers diminishing returns. The focus must now shift toward production-scale implementation and outcome-driven investments. She also pointed to leadership involvement as a major differentiator between organizations that successfully scale AI and those that remain stuck in experimentation mode. Unlike previous waves of digital transformation that could be delegated largely to IT departments, AI requires direct engagement from promoters, chief executives and boards because of its potential to reshape decision-making, operational processes and organizational structures. According to Kannan, many Indian enterprises continue to treat AI as a technology initiative rather than a business transformation exercise, creating a disconnect between strategic ambitions and operational execution. "Promoters, CEOs and boards need to take ownership and drive it almost like it is their own baby," she said. "They cannot simply delegate it and expect results." Data readiness The discussion further highlighted data readiness as a challenge. While AI models continue to become more powerful and accessible, many enterprises still struggle with fragmented systems, inconsistent data quality and unclear ownership structures. Ohmna Sinha, Global Head of Data & Analytics Governance at Nielsen, said these issues remain among the biggest reasons AI initiatives fail to progress from controlled pilot environments into large-scale production deployments. According to Sinha, organizations often underestimate the complexity of preparing enterprise data for AI systems. While data can be cleaned and standardized for a proof of concept, maintaining quality and consistency across an entire organization is significantly more difficult. She noted that many enterprises are still grappling with a basic governance question: everyone wants access to data, but few organizations have clearly defined who owns it. "One of the strongest contributors to failed proof-of-concepts is that organizations do not have clean data," Sinha said. "Data quality is not a one-time activity. As long as the model consumes data, the data has to remain accurate." Organisational learning Beyond data quality, she identified awareness and organizational learning as emerging barriers. In many enterprises, AI expertise remains concentrated within small specialist teams, while the rest of the workforce struggles to keep pace with rapid technological changes. This creates knowledge gaps that can slow adoption and reduce confidence in AI-driven decision-making. As AI continues to evolve, Sinha believes organizations must embrace continuous learning, encouraging employees not only to acquire new skills but also to unlearn outdated practices and relearn new ways of working. The rise of agentic AI and increasingly autonomous systems is adding another layer of complexity. While enterprises are eager to explore AI agents capable of independently performing business tasks, Sinha cautioned that governance cannot be treated as an afterthought. Human oversight, she argued, remains essential because issues such as bias, hallucinations and inaccurate outputs continue to pose risks in enterprise environments. Organizations may be moving toward greater automation, but they are not yet at a stage where critical decisions can be left entirely to autonomous systems. The consensus was that India's AI challenge has fundamentally changed. The question is no longer whether enterprises are willing to adopt AI, as most already have. The bigger question is whether they possess the discipline, governance structures, leadership commitment and data foundations necessary to turn AI investments into productivity gains and competitive advantage. For India, has built one of the world's largest pools of AI talent, hosts a rapidly expanding startup ecosystem and is investing heavily in national AI infrastructure. Yet, as Kannan observed, India's AI ambitions will ultimately be realized not through announcements, pilots or experimentation, but through the ability of enterprises to embed AI into the way they operate, govern and make decisions. AI may be generating excitement today, but the next phase of India's AI story will be determined by execution.