Trusted AI at Scale: A Global South Leadership Dialogue

Abstract

The round‑table brought together senior ICT leaders from the Global South to translate the summit’s “Safe & Trusted AI” agenda into actionable steps. Participants examined three thematic pillars – automotive AI safety, government‑industry‑academia coordination, and regulatory comparative analysis across major jurisdictions – with a focus on how India and France can jointly pioneer interoperable, trustworthy AI at scale. The discussion produced concrete recommendations for joint governance, talent pipelines, sandbox‑based testing, and collaborative pilots, underscoring the complementary strengths of India’s talent pool and France’s regulatory and infrastructure framework.

Detailed Summary

Thomas (Moderator) opened the session by highlighting the need to move from high‑level “Safe & Trusted AI” principles to concrete implementation pathways, especially for fast‑moving sectors such as automotive.

1.1 Automotive AI – Misconceptions & Technical Realities

A speaker (later identified as Kip Wainscott) outlined three core obstacles in deploying trustworthy AI in the automotive domain:

Misconception that Large Language Models (LLMs) = AI – LLMs receive massive funding and media attention, but they perform probabilistic anticipation rather than formal reasoning. They excel at user‑interface tasks (e.g., voice assistants) but are insufficient for safety‑critical verification and validation.
Role of Perception Models – For safety‑critical functions, perception models (sensor‑fusion, computer‑vision pipelines) provide deterministic analysis and can surpass human performance in data‑intensive decision‑making.
System Architecture – Modern cars employ a hierarchy of semi‑specialized agents (e.g., perception, planning, control) coordinated by a master orchestrator. This “agentic AI” architecture demands rigorous integration testing and verification.

Additional constraints discussed:

Legacy hardware in Indian and French factories hampers Industry 4.0 interoperability.
Data privacy & sovereignty – OEMs are reluctant to store proprietary R&D data on non‑sovereign clouds.

The speaker concluded that India–France collaboration could combine complementary capabilities (India’s scale, France’s regulatory rigor) to overcome these barriers.

2. Government Recommendations (India‑France AI Initiative)

Sarita (government representative) presented a three‑point agenda for public‑sector action, followed by Neha (industry lead) who elaborated on implementation pathways.

2.1 Launch an India‑France AI Initiative Mission

Purpose: Create a joint governance body that leverages India’s massive talent pool of AI & data engineers and France’s strong regulatory framework and “AI‑safety‑first” culture.
Outcome: Accelerate pilot‑to‑production pipelines, allowing India’s large‑scale deployment environment to serve as a test‑bed for trustworthy AI systems.

2.2 Data & Regulation Alignment

Problem: Data cannot move freely across borders without safeguards.
Recommendation: Co‑develop cross‑border data‑sharing protocols that preserve privacy while enabling AI training and inference across the two nations.

2.3 Industry & Academia Enablement (Neha’s Points)

Re‑position Global Capability Centers (GCCs) – Treat GCCs in India as innovation and co‑creation hubs, not mere back‑office units.
Industrial Metaverse & Digital Twins – Propose a “Katina X”‑style initiative (a joint industrial‑metaverse platform) to standardize data exchange between French and Indian factories.
Up‑stream Co‑creation – Embed AI quality gates early (OEM → Tier‑1 → Tier‑2) to accelerate time‑to‑market and boost innovation.
Talent Pipeline – Establish exchange programs, post‑doc fellowships, and joint curricula between French and Indian R&D institutes to create “Day‑One AI‑ready” talent.

2.4 Closing Government Perspective

Sarita emphasized that AI is not a silver bullet, but it is a real, actionable technology. The joint mission must focus on ownership of technology (rather than mere licensing) to position the Indo‑French partnership as a credible “third pole” in global AI geopolitics.

3. Comparative Regulatory Landscape (Legal Scholars)

Ahmed Baladi and Lorien Anceny (partners from Gibson Dunn and A&O Sherman) delivered a concise but thorough comparative study of AI regulation across major jurisdictions, aiming to answer the question: “If you had to develop an AI solution, where would you go?”

3.1 European Union – The AI Act

Structure: Traffic‑light system (Red = prohibited, Yellow = strict compliance, Green = minimal rules).
Complexity & Cost: Hundreds of pages, with an expected cascade of 100+ guidelines (paralleling GDPR). Compliance costs are high, and the time to certify may erode competitive advantage, especially for fast‑moving sectors.

3.2 India – Principle‑Based Flexibility

Approach: Leverages existing statutes (consumer protection, data protection) and principles rather than a monolithic AI law.
Key Principles (Nov 2025 Guidelines):
1. Build trust
2. Put people first
3. Innovate boldly
4. Be fair
5. Be accountable
6. Make AI understandable
7. Keep it safe
Outcome: Pragmatic, innovation‑friendly environment that can rapidly iterate while still safeguarding users.

3.3 United States – Deregulation Trend

Policy: Executive order (Trump era) emphasizing minimal federal regulation; states (e.g., California, Colorado) are developing patchwork AI statutes.
Implication: Regulatory uncertainty at the federal level but potentially lighter compliance burden for innovators.

3.4 China – Safety‑Centric, National‑Interest Model

Focus: Specific AI use‑cases (generative AI, recommendation systems, deep‑fakes) are regulated to protect national security and public order.

3.5 Positioning France & India

France – Must comply with the EU AI Act but is actively promoting a “sandbox” ecosystem (national data‑center policy) to attract AI investment and enable rapid prototyping.
India – Uses principle‑based, flexible regulation to harness its vast talent pool and large‑scale public‑sector AI platforms.

Key Insight: The regulatory regime shapes capital flows, talent migration, and the locus of breakthrough AI. No single model is universally optimal; the Indo‑French partnership can blend the EU’s rigor with India’s agility.

4. Concrete Franco‑Indian Collaboration Blueprint

Laurie (speaker) moved the conversation from analysis to actionable recommendations.

4.1 Strategic Rationale

India: World’s largest AI talent pool, strong engineering education, proven capability to run massive public‑sector AI platforms.
France: Europe’s AI leader with €109 bn in AI‑related investments, national data‑center policy facilitating rapid scaling of AI infrastructure.

4.2 Four Core Principles for Joint Governance

Non‑discrimination
Responsible use
Transparency
Human oversight

These principles are high‑level enough to be adopted across differing legal systems while providing clear direction.

4.3 Enabling Drivers

Driver	Description	Expected Impact
Joint Sandboxes	Controlled environments where French and Indian regulators test AI systems together before market launch. Analogous to flight‑simulator testing for aircraft.	Reduces regulatory surprises, builds mutual trust, accelerates learning.
Research Collaboration	Joint funding of AI research, exchange of PhDs, co‑development labs (e.g., Tata Consulting Services in France, Schneider Electric’s hub in India).	Ensures evidence‑based policy, aligns innovation with regulatory goals.

4.4 Implementation Pathways

Pilot Projects: Start with joint pilots in sectors already covered by the white paper (healthcare, automotive, governance).
Regulatory Alignment without Uniformity: Create a “regulatory alignment” framework that allows each country to retain its regulatory nuances while achieving interoperable compliance for multinational AI products.
Leverage Trade Agreement: The recent EU‑India trade agreement provides a diplomatic foundation for deeper AI cooperation.

4.5 Anticipated Benefits

Attract foreign investment by offering a predictable, dual‑jurisdiction compliance model.
Position the Indo‑French bloc as a third pole capable of influencing global AI standards.
Create a scalable model that other nations can join, expanding the alliance beyond the two founding members.

5. Closing Synthesis & Call to Action

Pratush Kumar (CEO, Savam AI) delivered the concluding remarks.

He highlighted the “inflection point” where AI moves from hype to an integral societal infrastructure—touching sovereignty, cross‑border collaboration, and human flourishing.
The white paper (year‑long effort, ~60 contributors, 23 young leaders) was praised for its breadth (healthcare, automotive, regulation) but emphasized that implementation velocity is the real test.
France’s early sovereign‑AI investments and India’s rapid home‑grown model building were identified as complementary strengths.
He called for real joint pilots, expanded sector coverage (finance, education, agriculture), and robust metrics to track progress.

The session wrapped with a group photo and a thank‑you to all participants, signaling the transition to the next agenda item.

Key Takeaways

LLMs ≠ Full AI – Large language models are useful for UI/UX but cannot replace formal reasoning or safety‑critical perception models in sectors like automotive.
India‑France AI Initiative Mission – A joint governance body that combines India’s talent depth with France’s regulatory rigor can accelerate trustworthy AI deployment.
Data‑Sharing Protocols – Cross‑border data flows must be secured through jointly crafted regulations to protect sovereign R&D data.
Industrial Metaverse & Digital Twins – A Franco‑Indian “Katina X” platform could standardize interoperability between factories, enabling a shared industrial metaverse.
Regulatory Comparative Insight –
- EU AI Act – High compliance cost, strict classification system.
- India – Principle‑based, flexible, innovation‑friendly.
- US – Deregulation trend, fragmented state‑level rules.
- China – Safety‑and‑national‑interest focused.
Joint Sandboxes & Research – Sandbox environments and coordinated research are essential “drivers” for aligning divergent regulatory regimes without forcing uniformity.
Four Core Governance Principles – Non‑discrimination, responsible use, transparency, and human oversight provide a universal foundation for Indo‑French AI collaboration.
Strategic Benefits – The partnership can become a global AI “third pole,” offering predictable compliance, attracting investment, and setting a template for other nations.
Implementation Over Documentation – The white paper’s recommendations must translate into pilot projects, talent‑exchange programs, and measurable milestones to realize impact.
Future Expansion – Beyond healthcare and automotive, the alliance should explore finance, education, and agriculture, using the existing Young Leaders Program as a catalyst.

See Also:

India AI Impact Summit 2026

Explorer

trusted-ai-at-scale-a-global-south-leadership-dialogue