C:\>DIR Global ‘Agentic Regulator’ Hackathon

Decentralised Market Infrastructure, Smart Contracts & AI Agents

AI agents now can transact directly with self-executing smart contracts - routing orders, managing collateral and rebalancing liquidity across DeFi and tokenised markets.

What agentic solutions can help public authorities to understand and monitor agentic on-chain activities and mitigate potential systemic vulnerabilities? 

Context for the Problem Space

AI agents now transact directly with smart contracts at scale, autonomously routing orders, managing collateral, and rebalancing liquidity constantly.

Risks include but are not limited to: 

  • Cascading liquidations triggered by agents acting simultaneously

  • Agent-layer vulnerabilities, including prompt injections and compromised wallets

  • Accountability gaps as autonomous agents lack legal identity under current regulations

  • Agentic biases: What form of digital money will agents prefer? Will they take into account the relative risks and constraints of traditional rails, stablecoins, Tokenised Deposits or CBDC?

Potential Solution Areas

Securing Agent-Layer Attack Surfaces

How can regulators protect digital assets against prompt injection and mandate exploitation at the largely invisible agent layer?

Bridging Accountability and Perimeter Gaps

How can enforcement agencies assign legal liability across jurisdictions when autonomous agents - lacking legal personhood - cause harm?

Establishing Know Your Agent (KYA) Standards

How can authorities implement verifiable identity frameworks to ensure autonomous agents operate strictly within authorized bounds before transacting?

Restoring Oversight Visibility

How can supervisors transition from periodic reporting to continuous, real-time monitoring of machine-speed transactions across decentralized networks?

Illustrative Hackathon Prototypes

Agentic Payments Observatory

Build a dashboard for agent-to-service or agent-to-agent payments, linking smart contracts and transactions to agent identity, permissions, receipts, anomaly detection, and policy breaches. 

Perimeter and Rulebook Copilot

A research agent that monitors DeFi incident data and governance to identify exactly where autonomous agents evade entity-based regulatory frameworks like MiCA.

KYA Maturity Assessor

An evaluation tool that scores protocols against "Know Your Agent" standards (verifiable identity, mandate limits, injection resilience) to generate standardized regulatory ratings.

Programmable Guardrails

Build a smart-account or contract layer that limits what a financial agent can do through permissions, spending caps, whitelists, approvals, session keys, and human escalation

Reputation Registry

Build an ERC-8004 registry for managing agent reputation.

Reference & Resources

Contact

For additional questions, please contact: contact@cdir.global