India-Japan AI for the World

Detailed Summary

Speaker: Shiho Nagano, Director for Information Policy Planning, METI

• AI Industrial Development: Japan is executing a comprehensive AI strategy that integrates foundation models, domain‑specific models, and bespoke corporate AI systems. The strategy leverages Japan’s rich industrial data to drive innovation across manufacturing, services, and other sectors.
• Computing Infrastructure: Emphasis on strengthening national data‑center capacity, semiconductor supply‑chain resilience, and high‑performance computing projects such as ABCI and collaborations with what‑bit (likely “WIDE/NVIDIA” mis‑heard).
• Frontier AI & AGI: Japan aims to nurture competitive foundation models and promote community collaboration (open‑weight sharing, data‑set exchange).
• GENIAC – Generative AI Accelerator Challenge (launched Feb 2024):
  1. Compute subsidies (GPU/cloud credits).
  2. Collaboration facilitation among developers, industry users, and data‑holders (events, matchmaking).
  3. Developer‑community building (knowledge‑exchange, networking).
• AI Governance Milestones:
  • 2023: Hiroshima AI Process at the G7 summit → policy framework.
  • Feb 2024: AI Safety Institute (via IT Promotion Agency).
  • Apr 2021: Publication of AI Business Guidelines.
  • Jun 2025: Enactment of the AI Act, positioning Japan as a “most innovation‑friendly” jurisdiction while ensuring safety, transparency, and alignment with global norms.
• Closing Remark: Invited participants to discuss cutting‑edge AI tech and challenges, aiming to deepen strategic ties between India and Japan.

Key Insight – Japan: A strong, government‑backed ecosystem that couples policy, compute resources, and a focus on trustworthy, sovereign AI.

---

2. Photo Session

• The moderator announced a group photo with six panelists:
  1. Mr. Sira Madan (BCG) – not on the provided speaker list.
  2. Mr. Shingo Okuma (Highreso) – confirmed panelist.
  3. Mr. Satish Thiagarajan (TCS Japan) – confirmed panelist.
  4. Mr. Lucas Haywood (OneStraction) – confirmed panelist.
  5. Mr. Sunil Gupta (Yotta) – confirmed panelist.
  6. Mr. Hirotaro Ohira (Fujitsu Research of India) – confirmed panelist.

The photo took several minutes; the moderator then asked the panelists to be seated.

---

3. Opening Keynote – India

Speaker: Golab Jha, India AI (representing the India AI Mission)

• Strategic Importance of AI for India: AI is a driver of productivity, complementing India’s historic strength in IT services.
• AI Vibrancy Index: India rose from 7th (2023) to 3rd (2025) after launching the India AI Mission (Mar 2024).
• Four Pillars for Future Growth:
  1. Foundation Models tailored to Indian languages & contexts (instead of importing western models).
  2. Data Availability: Consolidating fragmented datasets into a unified, accessible resource.
  3. Compute Access: Subsidised GPU/compute resources for researchers and startups.
  4. Responsible AI: Emphasising safety, ethics, and governance.
• India AI Mission Initiatives:
  • AI‑Kosh: Dataset marketplace – >9,000 datasets, free compute credits, toolkits, open‑source models.
  • Compute Pillar: >38,000 GPUs available with ~40 % subsidy.
  • Innovation Center: Indigenous AI component development.
  • Safe & Trusted AI Pillar: Ethical AI frameworks.
  • Future‑Skills Pillar: Upskilling across the AI value chain (data annotation to PhD research).
  • Startup‑Financing Pillar: Financial & non‑financial support (accelerators, go‑to‑market help).
  • Application Pillar: Government‑sourced problem statements, hackathons, and challenge programmes.
• Call to Action: Attendees invited to visit the India AI exhibit at the Bharat Pavilion and explore the AI‑Kosh demo.

Key Insight – India: A rapid‑scale, talent‑rich ecosystem backed by substantial government subsidies and a clear focus on building sovereign, responsible AI.

---

4. Panel Discussion – “Strengths & Collaboration for Sovereign AI”

Moderator: Takumi Miyakawa (METI) – kept time (3 minutes per panelist) and opened the floor.

4.1. Hirotaro Ohira – Fujitsu (India)

• Sovereign AI Definition: Trust, long‑term commitment, and disciplined engineering built over decades.
• Fujitsu’s Strengths:
  • End‑to‑end stack – secure AI models, energy‑efficient and sustainable compute, hybrid & quantum computing, networking.
  • Close partnership with NVIDIA for domain‑specific, secure AI hardware.
• Talent Profile: Fujitsu Research of India hosts 400 researchers (≈ 27 % of Fujitsu’s global research headcount). The centre draws talent from IITs, ISCs, and industry hires.
• Bridge Programme: Participation in Japan’s Lotus Programme (Japanese‑government‑funded student exchange) to bring Indian researchers to Japan.
• Collaboration Outlook: Japan offers trust & engineering discipline, India supplies scale & talent. Both sides fill each other’s gaps.

4.2. Sunil Gupta – Yotta

• India’s Dual Role – Supply & Demand:
  • Demand: Massive user base (≈ 1 billion smartphone‑enabled internet users) creates a “digital savviness” market.
  • Supply: Rapid adoption of AI once technology matures and costs fall (parallels to 4G/5G adoption).
• Frugal AI Economy: Emphasis on voice‑first services for non‑English speakers; potential to reach the global south and APEC markets.
• Synergy Idea: Combine India’s scale & market with Japan’s industrial‑automation expertise (Kaizen, lean manufacturing) to deliver AI‑driven factories and services globally.

4.3. Lucas Haywood – OneStraction

• Puzzle‑Piece Analogy:
  • Both countries are middle powers facing a polarized world; collaboration is strategically important.
  • Japan: Deep domain expertise (manufacturing, construction, standards).
  • India: Agility, speed, abundant IT talent.
• Start‑up Role: Agile start‑ups act as connectors between standards bodies, enterprises, and niche domains, accelerating AI adoption.
• Future Focus: Leverage start‑up agility to combine Japanese standards with Indian execution speed.

4.4. Satish Thiagarajan – TCS Japan

• Challenge – Knowledge Capture: Japan’s aging population stores expertise in heads, not data sets. Need to extract & codify this knowledge (e.g., via knowledge‑graphs).
• India’s Assets: Large talent pool, massive datasets (Aadhaar, UPI, government data) that can train sovereign models.
• Sovereign AI Vision: Not about isolation from hyperscalers but building nationally owned AI assets for country‑specific problems.
• Collaboration Suggestion: Joint development of knowledge‑graphs and domain‑specific models that leverage Japanese expertise and Indian data.

4.5. Shingo Okuma – Highreso (GPU Cloud Provider)

• Strengths of Japanese Data: High‑quality, well‑curated data from manufacturing and healthcare sectors.
• Indian Strength – Speed & Scale: Indian firms invest aggressively; Japanese firms tend to be more conservative.
• Collaboration Path: Highreso aims to expand GPU‑cloud services to India, needing local data‑center partnership and government support.
• Potential Offerings: Transfer of unique Indian AI platforms to Japan, co‑development of serving‑layer infrastructure.

4.6. “Mithi” (unidentified speaker) – Brief Comment

• Highlighted cultural differences in work style; pointed out the need for large‑scale acquisitions (especially data‑center assets) rather than piecemeal cooperation.
• Noted that Japanese enterprises are increasingly looking abroad for innovation and scale, especially in construction and manufacturing.

4.7. Moderator’s Closing

• Invited audience questions; emphasized time limitation (13 minutes) and moved directly to Q&A.

---

5. Audience Q&A

Questioner	Affiliation / Context	Core Question / Comment	Panelist(s) Responding
Aziz (Kyrgyz Republic)	Represents a talent‑pool country; copies Indian outsourcing model	Promoted an open‑source multilingual AI hub (≈ 7,000 languages) hosted on Hugging Face; asked about collaboration with India/Japan	No specific answer recorded; session ended shortly after.
Neeraj (Bangalore)	Former ML architect, now in education sector	Asked how academia‑industry partnerships can be nurtured; referenced a recent India‑Japan matching event; sought concrete cooperation ideas	Sunil Gupta (Yotta) outlined existing compute‑access programmes (GPU workstations, cloud credits) and suggested contacting the India AI Mission for funding and lab creation.
Unnamed attendee	Raised a cultural‑workforce gap question (Japanese labor shortage, need for Indian talent)	Asked whether large‑scale acquisitions (esp. data‑center or language‑model assets) could resolve the issue faster than incremental cooperation	Lucas Haywood (OneStraction) explained that Japanese firms are already acquiring Indian start‑ups and that a two‑way talent flow is emerging; large investments are expected to continue.
Additional brief comment	From a speaker with experience living in Japan (≈ 7 years)	Highlighted historical preference for Southeast‑Asian partners and urged deeper investment in India to overcome cultural reluctance	No direct answer; the moderator thanked the comment and moved to closing.

---

6. Closing & Announcements

• Panelists were invited to stand for a brief gratitude applause.
• Organisers announced that the Japan Pavilion (Hall 14) featured seven exhibiting companies, encouraging attendees to visit.
• Event formally closed with thanks to participants, audience, and sponsors.

Key Takeaways

• Japan’s AI strategy blends policy, compute infrastructure, and a focus on trustworthy, sovereign AI (GENIAC, AI Act).
• India’s AI mission delivers rapid scaling through massive data, subsidised compute, and talent development (AI‑Kosh, compute pillar, skill pillar).
• Complementary strengths:
  • Japan – engineering discipline, high‑quality industrial data, strong governance.
  • India – scale of market, abundant IT talent, frugal AI approaches for multilingual, low‑resource contexts.
• Collaboration opportunities identified:
  1. Joint knowledge‑graph projects to capture Japan’s legacy expertise.
  2. Co‑development of domain‑specific foundation models tuned to Indian languages and Japanese industry needs.
  3. Expansion of Japanese GPU‑cloud services (e.g., Highreso) into India with local data‑center partnerships.
  4. Start‑ups acting as connectors between Japanese standards and Indian execution speed.
  5. Large‑scale investments and acquisitions (data‑centers, language‑model assets) to accelerate sovereign AI ecosystems.
• Governance & Responsible AI remain central for both nations; both have launched national AI safety institutes and ethics guidelines.
• Talent exchange programmes (e.g., Japan’s Lotus Programme, India‑Japan university match‑making) are already operational and poised to expand.
• Audience engagement highlighted interest in multilingual open‑source AI, educational compute access, and the need for concrete pathways to turn policy commitments into investments.
• Future outlook: Panelists expressed optimism that the next five years could be a “golden period” for India‑Japan AI cooperation, driven by complementary capabilities and increasing geopolitical alignment.

---

See Also:

• indias-ai-infrastructure-from-vision-to-reality
• democratising-access-to-ai-through-data-infrastructure-models-governance-and-market-design
• financing-the-future-building-ai-ready-digital-foundations-for-asia-and-the-pacific
• ai-and-open-networks-creating-impact-at-scale
• from-guidelines-to-ground-institutional-ai-safety-for-india-and-the-global-south
• ai-for-all-role-of-open-source-hardware-and-software
• ai-impact-forum-democratising-ai-resources
• building-public-interest-ai-catalytic-funding-for-equitable-access-to-compute-resources
• sovereign-ai-infrastructure-for-bharat-and-global-south
• ai-innovators-exchange-accelerating-innovation-through-startup-and-industry-synergy