67% of Indian AI startups building apps: Report
India’s competition regulator has reported that nearly two‑thirds of domestic AI startups have focused on the application layer, and has urged self‑audits and transparency to curb risks.
India’s competition regulator has released a landmark market study showing that nearly two-thirds of domestic AI startups have focused on the application layer, a tilt that has begun to reshape how banks, hospitals, retailers and logistics firms deploy artificial intelligence across their operations.
The Competition Commission of India’s (CCI) study has also urged firms to conduct self‑audits of AI systems and adopt transparency measures to address emerging competition risks.
App-heavy focus on AI
Based on primary research, the CCI has reported that about 67% of respondent AI startups in India have been building applications, with roughly 20% operating in the data layer and around 10% in infrastructure.
The study has further noted widespread use of open‑source technologies and strong activity in machine learning, natural language processing, generative AI and computer vision.
The application-led startup mix has dovetailed with rapid adoption in end‑user sectors. The study has highlighted accelerating use of AI in BFSI, healthcare, retail and e‑commerce, logistics and marketing—covering use cases such as fraud detection, credit risk assessment, demand forecasting, dynamic pricing and personalised recommendations.
Open source and capability building
India’s startups have reportedly leaned on open‑source tooling (76% of respondents) to build and ship AI products faster and at lower cost, while focusing effort on domain‑specific data and models.
The CCI has said the study was conducted by Management Development Institute Society (MDIS), using interviews and surveys across startups, platforms, user sectors, investors and experts to capture ecosystem dynamics.
By the numbers
1) 67% of surveyed AI startups have been focused on application‑layer products; about 20% on data, and ~10% on infrastructure.
2) 76% have relied on open‑source technologies; activity spans ML (88%), NLP (78%), GenAI/LLMs (66%) and computer vision (27%).
The watchdog has flagged potential risks from ecosystem lock‑in, exclusive partnerships, opaque algorithms and AI‑facilitated collusion or price discrimination. It has recommended that enterprises conduct competition‑compliance self‑audits of AI systems, strengthen transparency around AI use in decision‑making, and participate in advocacy workshops.
The study has also proposed a dedicated think‑tank on digital markets and AI, and closer coordination with other regulators and international peers.
Building India‑first AI
The application focus has lined up with high‑profile efforts by Indian founders to localise AI. Bhavish Aggarwal—founder of Krutrim—has pursued an indigenous AI stack spanning models and infrastructure, underscoring the push to tailor systems for Indian languages and use cases.
In parallel, Sarvam AI co‑founder Vivek Raghavan has said the company has planned to open‑source models being trained under the IndiaAI Mission, signalling the sector’s openness to collaborative development.
CCI chairperson Ravneet Kaur has previously said AI has the potential both to aid fair competition and to facilitate collusion, reinforcing the case for algorithmic transparency and accountability.
The study has cited external estimates that the global AI market could expand from $244.22 billion in 2025 to $1 trillion by 2031, with India’s market rising from $7.84 billion to $31.94 billion over the same period—underscoring why competitive conditions in the country’s AI stack have mattered for growth and consumer welfare.


