India Rides the Agentic AI Wave with Over 80% of Businesses Exploring its Potential: Deloitte Report 70 per cent of firms indicated a strong desire to use GenAI for automation, highlighting the increasing adoption of AI-powered autonomous systems across industries.

Over 80 per cent of Indian organisations are actively exploring the development of autonomous agents, indicating a substantial shift towards Agentic AI, according to Deloitte's fourth wave of the State of GenAI report (India perspective).

The findings also highlight the growing interest in multi-agent workflows, which, with no constant oversight, orchestrates a set of goal-oriented sub-agents under the supervision of the master agents to perform the desired action, with 50 per cent of organisations identifying it as a key focus area.

Additionally, 70 per cent of firms indicated a strong desire to use GenAI for automation, highlighting the increasing adoption of AI-powered autonomous systems across industries. The survey further hints towards an increase in the pace of innovation, the democratisation of AI and investment in AI-driven transformation, with 61 per cent of firms actively pursuing more than 10 GenAI experiments.

For Indian businesses, the increased interest in GenAI and Agentic AI workflows is already producing a significant impact at scale. According to the report, more than 67 per cent of firms said that GenAI had a beneficial effect on all phases of the software development lifecycle.

The findings further demonstrated the business benefits of GenAI as almost 70 per cent of respondents said their AI integration efforts met or surpassed ROI estimates. Critical departments such as IT, customer service, marketing, operations, and product development emerged as leaders in AI adoption as businesses advance in their AI maturity.

"As Indian organisations explore Agentic and GenAI, the key to unlocking their potential lies in moving from experimentation to large-scale deployment. Businesses must build trust in AI systems by addressing concerns about errors, bias, and data quality through strong governance. Organisations must prepare to have a suitable ecosystem for the experiments to scale and be able to measure the impact that the use case drives. Our report indicates that most Indian organisations prefer buying AI solutions over developing them in-house. In such a case, ensuring adaptability to evolving needs is a challenge. Embracing an agile innovation approach is essential to stay ahead of AI advancements and optimise long-term returns. Balancing rapid adoption with sustainable strategies will be key to successful AI investment and growth," said Moumita Sarker, Partner, Deloitte India.

Despite the rapid adoption and enthusiasm around Agentic AI and GenAI, organisations face significant hurdles in scaling their initiatives. The survey found that while many businesses are experimenting with AI, only 29 per cent reported being able to fully scale up to 30 per cent of their AI proofs of concept, with the rest faring even lower. Additionally, AI adoption within workflows remains inconsistent, with 61 per cent of organisations reporting that only up to 40 per cent of employees with access to GenAI tools actively use them. Concerns around errors with real-world consequences (36 per cent), bias and hallucinations (30 per cent) and data quality (30 per cent) continue to slow down deployment. However, most organisations expect to overcome these challenges within 12–24 months, demonstrating confidence in AI's long-term potential.

One key roadblock to scaling AI is the approach organisations take towards adoption. The survey indicates that most businesses prefer buying GenAI tools and applications to developing them in-house. While this strategy allows companies to quickly integrate AI into their workflows, it may also lead to limited customisation, challenges in differentiation and dependence on external vendors. Organisations that fail to align AI solutions with their unique operational needs might struggle to realise the full potential of these technologies.

AI projects also have significant sunk costs, which complicates investment decisions. An overwhelming 94 per cent of firms would need more than six months to exit a project that does not achieve ROI goals, and 76 per cent expect it to take more than a year. This financial rigidity makes it essential for businesses to choose whether to build custom AI solutions or use pre-built tools right from the start. Given that most organisations prefer buying AI applications rather than developing them in-house, the challenge lies in ensuring these external solutions remain adaptable to changing needs.

This preference for off-the-shelf AI tools can increase the risk of obsolescence. With AI evolving at an unprecedented pace, 28 per cent of firms fear their current AI solutions will become outdated within the next two years. The reliance on third-party AI tools makes it critical for businesses to constantly evaluate whether their investments will stand the test of time or require costly replacements.