Multistakeholder Partnerships for Thriving AI Ecosystems

Detailed Summary

• Moderator (Robert Opp) opened the session by framing AI as a powerful lever for the Sustainable Development Goals (SDGs) while warning that, without responsible governance, the “AI equity gap” could widen existing inequalities.
• He referenced the Hamburg Declaration (first issued at the Hamburg Sustainability Conference) as a multi‑stakeholder pledge to embed Responsible AI in development work.
• The moderator announced a brief Q&A later and introduced the four panelists.

2. Government Perspective – Dr. Bärbel Kofler

2.1 The “Power Gap”

• Described AI’s promise for health, climate prediction, public‑service delivery and citizen‑government interaction.
• Emphasised that the real obstacle is not a lack of innovation but a power gap: venture‑capital, data‑centre capacity and skilled talent are heavily concentrated in the Global North.
• Cited statistics: only 17 % of global venture capital reaches the Global South, and 0.1 % of data‑centre capacity is located there.

2.2 Government as Enabler

• Outlined three levers governments must provide:
  1. Regulatory frameworks that guarantee privacy, ethical use and accountability.
  2. Infrastructure – reliable energy, compute resources and data‑sharing platforms.
  3. Human‑capital development – vocational and university programmes, and incentives for SMEs to adopt AI.
• Stressed the need for an open‑source, open‑data ecosystem to avoid AI benefits being captured by a few large players.

2.3 concrete commitments (Hamburg Declaration)

• Training target: 160 000 people in one year → achieved 190 000.
• AI building blocks for climate action: target 12 → delivered 15.
• Digital public‑goods datasets: target 30 → delivered 55.

2.4 International collaborations

• Mentioned pilots in Kenya (satellite data for farmers), Cambodia (cervical‑cancer detection), and multilingual data‑set work for Indian languages.

3. Private‑Sector Perspective – Arundhati Bhattacharya (Salesforce)

3.1 Democratising Technology

• Highlighted Salesforce’s “1 %‑model” (1 % of profit, product and employee time to community) and its legal requirement in India to dedicate 2 % of profits to social impact.

3.2 Financial‑Inclusion Case Study

• Described the PM Jandhan Yojana partnership with the State Bank of India: biometric UID‑AI enabled KYC, mobile‑network reach to 600 000 villages, and the creation of 97 % zero‑balance accounts that later received subsidies directly.
• The resulting digital cash‑flow enabled credit‑scoring for low‑income borrowers, reducing interest rates from 10 % per month to 7 % per year.

3.3 Skills and Community Building

• Salesforce has trained 3.9 million “Trailblazers” in India (second only to the U.S.).
• Emphasised that skill development is a prerequisite for AI adoption; without it, technology remains under‑utilised.

3.4 Role of Policy Makers

• Stated that adoption will happen automatically once technology demonstrably improves lives.
• Policy makers must ensure ethical safeguards, privacy protection, and infrastructure to avoid exploitation.

4. Mission‑Driven AI Start‑up – Nakul Jain (Wadhwani AI Global)

4.1 The “Technology is the Easy Part”

• Asserted that building AI models is straightforward; the real challenge lies in institutionalisation, governance, and field‑level support.

4.2 Education Use‑Case (Oral Reading Fluency)

• Partnered with the Government of Gujarat to embed an AI‑driven reading‑assessment tool within existing school workflows.
• Worked with a government technical partner to ensure data availability, annotation and teacher training, avoiding workflow disruption.

4.3 Healthcare Use‑Case (Tuberculosis)

• Collaborated with ICMR to define evaluation criteria from day‑one, ensuring that health outcomes are measured rigorously rather than as an afterthought.

4.4 Role as a Convener

• Described Wadhwani AI as a bridge that brings together government, NGOs, local implementers and technical experts, providing “hand‑holding” to keep pilots from stagnating in labs.

5. Corporate & Research Perspective – Dr. Sachin Loda (TCS)

5.1 Core Challenges

1. Data scarcity & fragmentation – most AI models rely on Western datasets, which are poorly suited to local contexts.
2. Compute asymmetry & energy – limited high‑performance hardware in the Global South; need for cheaper, repurposed legacy hardware and emerging quantum‑computing resources.
3. Talent gap – shortage of AI‑skilled professionals.

5.2 Open‑Data Ecosystem

• Advocated for a national sensing infrastructure (e.g., widespread air‑quality sensors) that feeds high‑velocity, high‑quality data into AI pipelines.
• TCS participates in AI Centres of Excellence (e.g., with IIT Kanpur) that focus on sustainability domains.

5.3 Compute Initiatives

• Mentioned a “Quantum Valley” partnership between TCS, IBM and the Andhra Pradesh government to explore next‑generation hardware.

5.4 Green AI & Evaluation Platform

• Introduced the “Trustee Platform” – an internal tool for evaluating model fairness, robustness and carbon footprint.
• Highlighted collaboration with Carnegie Mellon University on responsible‑AI research and standards.

6. Updating the Hamburg Declaration – Progress & Next Steps

Commitment (original)	Status (2024)
Train 160 000 people (AI skills)	190 000 trained (exceeds target)
Release 12 climate‑action AI building blocks	15 building blocks released
Publish 30 AI datasets as Digital Public Goods	55 datasets publicly available
Enable multi‑stakeholder pilots in Kenya, Cambodia, India	Ongoing pilots delivering satellite‑data for farmers (Kenya) and AI‑assisted cervical‑cancer screening (Cambodia)

• Dr. Kofler called for measurable, time‑bound commitments from all signatories and announced the next Hamburg Declaration summit (June 29‑13 2024).
• QR code displayed for attendees to endorse the declaration.

7. Emerging Opportunities & Suggested New Partnership Models

Speaker	Suggested New Model
Arundhati Bhattacharya	Expand national skilling missions (e.g., NASSCOM‑AICTE internships) to embed AI curricula in higher‑education and community colleges.
Nakul Jain	Create a global repository of reusable AI solutions (playbooks, code, governance templates) to enable startups from one country to deploy in another (e.g., India → Ethiopia).
Sachin Loda	Set up regional AI‑assurance hubs that provide standardized evaluation and certification for AI products, harmonising metrics across jurisdictions.
Dr. Kofler	Strengthen open‑data public‑goods pipelines and ensure government‑led data‑sharing policies that facilitate cross‑border AI research.

8. Q&A Highlights

Question	Key Points from Panelists
LLM vs. traditional ML for low‑resource settings (Nakul)

---

See Also:

• flipping-the-script-how-the-global-majority-can-recode-the-ai-economy
• thriving-with-ai-human-potential-skills-and-opportunity
• ai-innovators-exchange-accelerating-innovation-through-startup-and-industry-synergy
• human-flourishing-in-the-age-of-ai-a-global-south-coalition-for-transition-infrastructure-event-description
• shaping-resilient-futures-in-the-age-of-ai-leadership-for-the-technology-energy-and-security-transitions
• ai-and-the-new-frontier-for-economic-progress-linking-innovation-to-inclusive-growth
• catalyzing-global-investment-for-equitable-and-responsible-ai-in-health
• democratizing-ai-resources-and-building-inclusive-ai-solutions-for-india
• ai-for-inclusive-and-resilient-agricultural-food-systems
• mahaai-building-safe-secure-and-smart-governance