Mark Suzman: India must get four AI levers right in next 18–24 months for equitable gains

# YourStory Mark Suzman, CEO of the Gates Foundation, spoke to YourStory founder Shradha Sharma shortly before the 2026 Goalkeepers report launch. The foundation's report warns that artificial intelligence will either reduce or increase inequality depending on choices made now. To back the warning, the foundation committed at least $1 billion over two years to help bring AI tools to health, education and agriculture, with a focus on underserved communities.

Four practical priorities for India

Suzman identified four concrete things India must get right in the next 18–24 months if AI is to be an equalising force.

• Local-language access. Most early large language models were trained on English-dominant data. Suzman says AI systems must work in the languages people actually speak and be accessible as spoken language on everyday phones—so health workers, teachers, and smallholder farmers can use them directly.

• Build for local context. AI tools must be customised to specific crops, soils, irrigation practices, government schemes and state-level conditions. That requires locally relevant data and partnerships with institutions that understand regional variations.

• Data protections and governance. Suzman emphasised protecting personal data and national or sovereign datasets. Effective governance mechanisms must be in place before wide deployment so data use is transparent and rights are preserved.

• Invest in people and access. Training local workers to support AI tools is essential, along with improving access to compute resources. The foundation sees both skills and infrastructure as necessary complements to technology.

Examples seen on the ground

Suzman described recent visits to India that informed his views. He has visited India frequently for about 15 years and noted progress in health, education and agricultural development.

Two concrete case studies he cited:

• Eka Care: A partner in India that helps community health workers track pregnancy data, triage risk, and identify nutritional needs through an app that uploads and analyses client information.

• Wadhwani AI cough app: An application that analyses the sound of a cough recorded on a phone to screen for tuberculosis and determine whether further clinical triage is needed.

Suzman also mentioned partnerships with ICAR on climate-resilient crops and AI tools for more efficient farming.

Why the timeframe matters

Jobs and livelihoods

Suzman acknowledged AI's disruptive potential but framed it as a tool that can enhance roles in health, education and agriculture when deployed appropriately. Examples include enabling healthcare workers to triage more efficiently, helping teachers customise lesson plans, and supporting farmers with context-specific advice.

What this implies for stakeholders

For policymakers: create clear data governance, language and localisation strategies, and compute access plans.

For funders and philanthropies: target investments toward language models, contextual datasets, training programs and compute for underserved areas.

For private sector and developers: design tools for local conditions and languages, and partner with health, education and agricultural agencies to pilot and scale responsibly.