Start-ups and Global Capability Centers (GCCs) in India are leading firms in artificial intelligence (AI) adoption and are swiftly advancing from proof of concepts (POCs) to the production stage with Generative AI (GenAI), while domestic legacy enterprises are more cautiously evaluating the enterprise-grade functionality and reliability of the technology and are lagging behind, said a recent study released by EY on Thursday.

According to the report titled ‘Is Generative AI Beginning to Deliver on Its Promise in India? – AIdea of India Update’, just 15-20 per cent of POCs by domestic legacy enterprises have been rolled out into production, compared to GCCs which are rapidly embracing GenAI for innovation, having rolled out 30-40 per cent of POCs to production.

Further, the analysis of India’s top 50 most valued Indian unicorns in the study revealed that 66 per cent of them were already using AI or GenAI technology in their operations.

“This indicates that start-ups have been more proactive in terms of adopting AI technology as compared to the legacy players,” said the report.

The study highlights three issues – hallucination of Large Language Model (LLM) responses, data privacy and sovereignty, and the cost implications of deploying GenAI in production, as the key challenges for organizations in making investment decisions.

Talking about this, Mahesh Makhija, Partner and Technology Consulting Leader, EY India, said, “Enterprises need to make the shift from ad hoc experiments to deploying ‘fit for purpose’ use cases for immediate value creation and long-term programs that provide functional transformation. They will not only have to build their enterprise AI platform but also keep a close eye on cost implications/total cost of ownership and stay agile to adopt a hybrid approach due to the lack of India availability of different models.”

On GenAI use cases, the study said that about one-third of GenAI use cases involved using intelligent assistants for specific tasks, while about 25 per cent focused on marketing automation through text generation and text-to-image or text-to-video capabilities.

Another significant area, accounting for about 20 per cent of use cases, was document intelligence, which included document summarisation, enterprise knowledge management, and search.

“Additionally, companies are exploring GenAI for customer-facing chatbots, enhanced user experience, coding assistants, and internal process automation,” said the report.

First Published: May 16 2024 | 7:47 PM IST

Source link