India and Australia are making major moves to develop homegrown artificial intelligence systems, directing government funding and research partnerships toward building AI models they can control and trust for sensitive national applications. India's government has tapped two homegrown AI initiatives to develop cybersecurity models, while Australia released a comprehensive roadmap calling for sovereign AI capability in healthcare. Both nations are treating AI as strategic infrastructure, not just a technology to adopt.

Why Are Nations Building Their Own AI Systems?

The push for sovereign AI stems from growing concerns about relying on foreign AI models for sensitive government and national security applications. When a country depends entirely on international AI platforms for cybersecurity, healthcare, or defense, it surrenders control over how those systems work, what data they access, and how they're updated. India's government has explicitly stated it wants to reduce dependence on foreign frontier AI models in strategic sectors including cybersecurity, defense, governance, and public services.

Australia's healthcare leaders framed the challenge differently but with the same urgency. The nation's new AI roadmap emphasizes the need for "sufficient technological independence, local expertise, infrastructure and governance to reduce reliance on international technology providers". This isn't about rejecting global AI; it's about building local alternatives that can be tailored to national needs and governed within national borders.

What Are India's Specific Plans for Sovereign AI?

India's government has directed Sarvam AI and BharatGen, an IIT Bombay-led consortium, to develop indigenous AI models specifically for cybersecurity. These models will need to support cyber threat detection, malware analysis, vulnerability assessment, anomaly detection, and incident response across critical sectors including defense, banking, telecommunications, power, and government services.

The initiative operates under India's ₹10,300 crore IndiaAI Mission, a massive investment in building the country's sovereign AI ecosystem. The government has already onboarded more than 38,000 graphics processing units (GPUs), which are the specialized chips needed to train AI models, and is providing subsidized computing resources to startups, researchers, and academic institutions developing indigenous AI technologies.

Sarvam AI, founded in 2023, is building India's first indigenous foundation model designed for reasoning, voice capabilities, and multilingual applications, with dedicated government-backed computing infrastructure. BharatGen is developing multimodal foundation AI models capable of understanding text, speech, and images across more than 22 Indian languages, designed to support applications across governance, healthcare, education, and agriculture.

How Are Countries Implementing Sovereign AI Strategies?

• Infrastructure Investment: India has onboarded over 38,000 GPUs and is providing subsidized computing resources to startups and researchers, while Australia's roadmap calls for major investment in AI research through dedicated funding pathways and large-scale research centers.
• Sector-Specific Development: India is focusing on cybersecurity models for defense and critical infrastructure, while Australia is prioritizing healthcare AI with locally calibrated systems optimized to Australian populations and health needs.
• Workforce and Governance: Australia's roadmap recommends AI training for the healthcare workforce and creation of a national Chief AI Officer training program, while India's mission includes startup funding and research support for indigenous AI development.
• Privacy and Data Sovereignty: Australia emphasizes privacy-preserving technologies like federated learning, which enables AI development without sharing raw patient data across institutions, ensuring systems remain locally controlled.
• Multilingual and Cultural Adaptation: India's models are designed to preserve linguistic and cultural diversity across more than 22 Indian languages, addressing a gap that global AI models often miss.

Australia's approach includes a broader governance framework. The nation's National Policy Roadmap for AI in Healthcare, developed by the Australian Alliance for Artificial Intelligence in Healthcare and supported by multiple research institutions, makes 28 recommendations across eight priority areas. These include leadership, governance, sovereign capability, workforce development, consumer engagement, industry, and research.

"Healthcare is well known for its inertia to change and so we must find new agile ways to meet these historic challenges. Our mantra must be: Don't walk. Run," said Professor Enrico Coiera, lead author and founder of the Australian Alliance for Artificial Intelligence in Healthcare.

Professor Enrico Coiera, Lead Author and Founder, Australian Alliance for Artificial Intelligence in Healthcare

What Does Sovereign AI Mean for Patients and Citizens?

For healthcare, Australia's roadmap proposes a national AI-enabled "front door" for healthcare, using AI to improve access to and navigation across services such as Medicare and aged care programs. This means citizens could have a single AI entry point to help them navigate the health system, rather than struggling through multiple websites and phone lines.

For cybersecurity, India's push means that critical infrastructure like power grids, banking systems, and government networks would be protected by AI models developed and governed within India, reducing the risk of foreign surveillance or control. The models would be trained on Indian data and designed to understand threats specific to India's digital environment.

Both nations recognize that imported AI models, while powerful, may not serve local needs effectively. Global models are often trained on data that doesn't reflect local populations, languages, or contexts. By building sovereign AI, countries can ensure their systems are calibrated to local realities and governed by local rules.

"We have witnessed first-hand across our research projects and initiatives the increasing use of AI across health settings. The challenge now is to ensure adoption is guided by clear evidence, strong governance and collaboration across the sector," noted Adjunct Professor Annette Schmiede, CEO of the Digital Health Cooperative Research Centre.

Adjunct Professor Annette Schmiede, CEO, Digital Health Cooperative Research Centre

The broader context is clear: nations are treating AI as strategic infrastructure comparable to electricity grids or water systems. India's government has selected 12 organizations and research consortia, including Sarvam AI, BharatGen, Gnani AI, Fractal Analytics, and Tech Mahindra Maker's Lab, to develop large and small language models trained on Indian datasets. This diversified approach spreads the effort across multiple teams rather than relying on a single vendor.

Australia's roadmap was informed by extensive consultation across the healthcare sector, incorporating input from more than 96 organizations and 20 peak bodies and government agencies. This collaborative approach suggests that sovereign AI isn't about isolation; it's about building local capacity while maintaining the ability to collaborate internationally on terms that protect national interests.

As these initiatives unfold, they signal a fundamental shift in how nations approach AI adoption. Rather than asking "Which global AI platform should we use?", governments are now asking "How do we build AI systems that serve our unique needs and remain under our control?" The answer, for both India and Australia, involves sustained investment, local expertise, and a commitment to treating AI as essential national infrastructure.