The IndiaAI Mission, under the Ministry of Electronics and Information Technology (MeitY), has signed a memorandum of understanding with Karya, a nonprofit organisation that works to increase accessibility of the artificial intelligence economy. The collaboration is intended to strengthen India’s inclusive artificial intelligence ecosystem through coordinated work on data, technology and capacity building. The agreement emphasises shared responsibility for developing resources that serve diverse populations and languages. The memorandum focuses on the development, curation and sharing of high quality language and multimodal datasets to support inclusive and representative systems. Both parties will undertake technical cooperation to enhance the AIKosh data infrastructure, refine model evaluation frameworks and establish standards for dataset quality, validation and interoperability. The work is expected to create datasets that can be reused across research and policy applications. Ms Kavita Bhatia, Scientist G, GC, MeitY and chief operating officer of the IndiaAI Mission, described the collaboration as bringing together complementary strengths to build resources that reflect the full diversity of the population and its languages. She said that prioritising representativeness and quality in datasets is central to building a truly inclusive ecosystem. The statement indicated that the partnership will aim to align technical work with social needs. The pact includes initiatives for capacity building and ecosystem development through training programmes, workshops, technical consultations and knowledge sharing across government bodies and partner institutions. These activities are intended to support researchers, developers and policymakers in adopting standards and tools that foster inclusion. The partners expect that sustained engagement will improve accessibility of the artificial intelligence economy and the practical utility of datasets for diverse communities. The collaboration will also explore inclusive governance and ethical practices in dataset use.