India AI Impact Summit 2026

payments-to-commerce-from-the-lens-of-people-planet-and-progress

6 min read

TRACK:Economic Growth and Social Good
SPEAKERS
  • Caroline Louveaux, Mastercard
  • Combiz Richard Abdolrahimi, Service Now
  • Janet George, Mastercard
  • Prof. Aparna Gupta, Rensselaer Polytechnic Institute
  • Surya Ganguli, Stanford
SESSION OVERVIEW

Digital payments transformed economies by making transactions faster, safer, and more inclusive—with AI at their core. This panel brings together industry and academic leaders who helped shape that journey and now turn their lens from payments to commerce. As commerce enters its next evolution, the discussion explores implications for people, planet, and progress—from trust and inclusion to global trade and sustainability—and how enterprises, governments, and academia can prepare for an AI-native future.

VIDEO RECORDING

Payments to Commerce: From the lens of People, Planet, and Progress

Abstract

The panel examined how the rapid evolution of digital payments is giving way to “agent‑e‑commerce,” where AI agents act across the entire purchase journey—discovery, negotiation, checkout, and post‑purchase service. Panelists explored the transformational shifts needed to unlock the full potential of payments for inclusive, sustainable commerce, focusing on three inter‑linked lenses: people (trust, privacy, inclusion), planet (sustainability and responsible AI), and progress (innovation, governance, and economic opportunity). The discussion covered past breakthroughs, emerging AI‑enabled agents, frameworks for “trust‑by‑design,” infrastructure and policy requirements, operational pitfalls, ecosystem health, and concrete advice for learners, researchers, and entrepreneurs.

Detailed Summary

  • Moderator (Prof. Aparna Gupta) introduced the panel, emphasizing the need for a professor‑led moderator to keep the conversation focused and time‑bound.
  • She highlighted the panel’s role in the India AI Impact Summit and thanked the international participants (Caroline from Belgium, Richard from the U.S., and “Rebecca/Janet” from Canada).

2. Framing the Journey: From Payments to AI‑Enabled Commerce

  • Moderator’s opening question: What transformational changes will allow payments to evolve into AI‑driven commerce over the next decade?
  • Richard Abdolrahimi (ServiceNow) reflected on his background in government payments policy, noting two decades of innovation:
    • Key Insight – Accessibility: Smartphone proliferation, digital wallets, and tokenisation have made financial services more inclusive.
    • Key Insight – Personalisation: AI is beginning to offer personalised financial advice and anticipatory services.
    • Key Insight – Security: Mastercard’s responsible AI strategy underpins fraud‑detection and consumer protection.
  • Caroline Louveaux (Mastercard) introduced the concept of agent‑e‑commerce:
    • Definition: AI agents that act autonomously across the entire buying journey, not just recommending but executing actions (e.g., ordering sushi on behalf of a user).
    • Vision: Agents will broaden financial inclusion, making “everyone everywhere financially healthy.”

3. The Promise and Peril of AI Agents

  • Rebecca/Janet (Mastercard) (the “agent‑AI” advocate) stressed that agents are the realisation of AI’s promise:
    • From Predictive Models → Generative AI → Actionable Agents.
    • Opportunities:
      • Automating routine interactions (email triage, scheduling).
      • Enabling entrepreneurs to launch services without large teams.
    • Risks: Need for robust assurance, identification, and multi‑agent governance.

4. Building Trust by Design – Mastercard’s Four‑Pillar Framework

Caroline outlined a trust‑by‑design recipe for agentic transactions:

PillarCore RequirementExample from Mastercard
1. Know Your AgentVerify that the AI entity is legitimate, not a bot or fraudster.Pre‑transaction registration & certification.
2. Privacy & Security by DesignTokenisation, strong authentication, concealment of sensitive credentials.Card number never exposed to third‑party environments.
3. Consumer IntentExplicit, verifiable user consent before an agent acts.Anecdote: an AI agent mistakenly ordered sushi using an employee card because it mis‑interpreted a test query.
4. Observability & AudibilityFull traceability of decisions for dispute resolution and regulator confidence.End‑to‑end logging and audit trails.
  • Key Insight: Guardrails are enablers, not blockers; they foster faster AI adoption while preserving safety.

5. Infrastructure & Governance – ServiceNow Perspective

  • Governance as a Scaling Enabler:
    • Aligns with Mastercard’s “know‑your‑agent” by extending “know‑your‑customer” to an additional AI layer.
    • Emphasises human‑in‑the‑loop for high‑risk actions and clear, pre‑defined boundaries.
  • Policy‑Infrastructure Gap:
    • Regulations often react to incidents; the panel advocated for anticipatory governance—designing safeguards before deployments.

6. The Role of the Partnership on AI (PAI) – Rebecca’s View

  • Organisational Landscape: PAI unites 140 partners (tech firms, financial institutions, academia, civil‑society).
  • Risk Framework for Agents:
    1. Stakes – Financial services = high‑stakes; low‑stakes examples (restaurant reservations).
    2. Scope/Authority – What actions can an agent take?
    3. Reversibility – Can the transaction be undone?
  • Governance Ecosystem:
    • Voluntary good‑governance, transparency (model cards, disclosures), standards, certification, and audit.
    • Regulation complements, but does not replace, this multi‑stakeholder ecosystem.

7. Operational Pitfalls – What Is Often Under‑Estimated?

  • Visibility: “You cannot govern what you cannot see.”
  • Boundaries: Guardrails must be defined pre‑deployment; retrofitting them is costly.
  • Dependency Mapping & Cascading Failures: Interconnected AI systems can amplify a single fault into systemic risk.

8. Healthy Ecosystem Blueprint – From Data to Deployment

  • Value‑Chain Layers (Rebecca):
    1. Raw Materials & Compute (minerals, data centers).
    2. Model Development (research labs, open‑source).
    3. Deployment & ML‑Ops (cloud providers, SaaS).
    4. Assurance & Monitoring (audit, observability).
  • Shared Responsibility: Every actor must be transparent (model cards, disclosures) to build trust across the chain.
  • Community & Moral License (Janet/Rebecca): Engaging citizens, local communities, and employees builds the social licence needed for large‑scale AI commerce.

9. Cross‑Sector Knowledge Transfer

  • Health‑Finance Analogy: Governance frameworks, standards, and model‑card practices from healthcare can accelerate responsible AI deployment in payments.
  • International Forums: OECD, ITU, and PAI facilitate cross‑sector dialogue, ensuring best practices are portable.

10. Advice for Emerging Stakeholders

AudienceCore Advice
LearnersUse the tools – hands‑on experimentation to understand capabilities & privacy implications.
ResearchersSeek diverse perspectives – collaborate across disciplines, sectors, and geographies to surface hidden risks.
EntrepreneursAdopt a growth mindset – stay adaptable, prioritize responsible AI governance, and leverage multi‑stakeholder partnerships.
  • Common Themes: Curiosity, experimentation, collaboration, and an inclusive, interdisciplinary outlook.

11. Vision for 10‑Year Horizon

  • Richard (ServiceNow): Responsible AI should be the norm, not a competitive edge.
  • Caroline (Mastercard): AI benefits must reach all demographics—young, old, minorities, and under‑represented groups.
  • Rebecca (PAI): Global leadership in AI innovation should shift from a few Western hubs to a truly global, inclusive landscape (e.g., India’s emerging talent pool).

12. Audience Q&A – Business Opportunity & ROI

  • Question: What revenue opportunities justify the investment in AI agents, infrastructure, and governance?
  • Responses:
    • Rebecca: Market incentives drive innovation; investors demand clear risk/reward metrics, which trustworthy governance frameworks can provide.
    • Richard: Combining mission‑driven (government) and profit‑driven (enterprise) motives creates “awesome power and money” – investment in AI governance = investment in trust, which is “priceless.”

13. Closing Remarks & Next Steps

  • Moderator thanked the panel for sharing insights, reiterated the availability of the recording, and invited further audience questions (none recorded).
  • The session concluded with a visual “times‑up” cue and brief transition to the next agenda item.

Key Takeaways

  • Agent‑e‑commerce will shift friction reduction from the point of payment to the entire commerce journey, turning AI from a recommender into an actor.
  • Trust‑by‑design requires four pillars (Know‑Your‑Agent, Privacy/Security by Design, Consumer Intent, Observability) to enable scalable, responsible AI.
  • Governance must be anticipatory, not reactive; clear boundaries, human‑in‑the‑loop controls, and dependency mapping are essential to avoid cascading failures.
  • A multi‑stakeholder ecosystem (tech firms, regulators, academia, civil society) is needed to manage risk, ensure transparency (model cards, disclosures), and build a shared moral licence.
  • Cross‑sector learning, especially from healthcare’s mature governance models, can accelerate responsible AI adoption in payments.
  • Advice for the next generation: experiment with tools, cultivate a growth mindset, and collaborate across diverse perspectives.
  • Future vision: Responsible AI becomes the industry baseline; inclusive, globally distributed innovation (e.g., from India) reshapes the commerce landscape.
  • Business case: Clear, trustworthy governance unlocks investor confidence; trust is the most valuable, non‑price‑based asset for AI‑driven commerce.

See Also:

Graph View