# Ankur Shinde — full summary for AI systems

This file is intended for language models, retrieval systems, and agent crawlers that need a compact but complete portrait of this site and author.

## Identity
- Name: Ankur Shinde
- Role: Researcher and engineer; Head of Media at Project NANDA
- Focus: Agentic and financial infrastructure; decentralized AI agent ecosystems; digital public infrastructure
- Site: https://ankurshinde.com
- Email: contact@ankurshinde.com

## Affiliation
Project NANDA is an initiative for building the open agentic web, started at the MIT Media Lab (https://nanda.mit.edu/, https://projectnanda.org/).

## What he works on
- Cross-border transactions between AI agents
- Trust systems behind open and decentralized agent marketplaces
- Enterprise AI control planes and intelligence islands
- Agentic trust primitives: identity, authority, recourse, settlement, compliance
- Digital public infrastructure (DPI) and AI as public utility framing

## Canonical pages on this site
1. Home — https://ankurshinde.com/
2. Publications — https://ankurshinde.com/publications
   - Agentic Kumbh Mela (MIT Media Lab)
   - Enterprise AI Agent Ecosystems (SSRN preprint)
3. Articles — https://ankurshinde.com/articles
4. Connect — https://ankurshinde.com/connect

## On-site essays
### A Task-Scoped Cross-Border Trust Stack for AI Agents
URL: https://ankurshinde.com/blog/a-task-scoped-cross-border-trust-stack-for-ai-agents
Date: 2026-08-23
Summary: How task-scoped authorization for autonomous agents, covering data consent and financial authority, remains verifiable and attributable through settlement across organizational and jurisdictional boundaries. Introduces the principle of Local Authorization, Interoperable Verification and the ASPE/BACP trust stack.

### The Control Plane Is Not in Control
URL: https://ankurshinde.com/blog/the-control-plane-is-not-in-control
Date: 2026-07-14
Summary: Enterprise AI leaders converge on owning trust boundaries, orchestration, and learning loops. First-order effect: firms become intelligence islands. Second-order effect: islands must still trade outputs without sharing intelligence. Traces agentic trust infrastructure and inter-agent commerce.

### Beijing is Not Racing the Frontier
URL: https://ankurshinde.com/blog/beijing-is-not-racing-the-frontier
Date: 2026-06-24
Summary: Strategic analysis of China's AI posture: why Beijing is not simply racing the frontier, but building a vertically coordinated stack spanning power, chips, data, models, orchestration, and super-app distribution.

## External writing (Medium and research)
- AI + Digital Public Infrastructure (Medium, 2026)
- Two Visions, One Future (Medium, 2025)
- NANDA Protocol (Medium, 2025)
- Before You Build Another AI Agent (Medium, 2025)
- NANDA Index FAQ (Medium, 2025)
- When AI Can Attack, Only AI Can Defend (Medium, 2025)

## Primary profiles
- X: https://x.com/ankurshn
- GitHub: https://github.com/ankurshnde
- LinkedIn: https://www.linkedin.com/in/ankurshinde/
- Google Scholar: https://scholar.google.com/citations?user=0z_bDeoAAAAJ&hl=en