Proposal: Daml AI Security Tooling and Sherlock AI ecosystem-wide coverage

Development Fund Proposal Submission

Proposal file: proposals/sherlock-ai-canton.md

---

Summary

Sherlock will customize its AI auditing engine for Daml and Canton's architecture, then provide 25 AI audit runs to projects building on Canton. Each run includes a full AI-driven security review with human triage from researchers with Canton codebase experience. ---

Checklist

• [x] Proposal file added under /proposals/
• [x] Milestones and funding amounts defined
• [x] Acceptance criteria included
• [x] Alignment with Canton priorities described

---

Notes for Reviewers

• If both this proposal and the contest audit proposal are approved, they create a flywheel: contest findings become direct training data for the AI, and contest researchers staff the human triage layer
• The AI customization is a one-time investment that creates a permanent capability for the Canton ecosystem, improving over time as more codebases are analyzed
• The 25 runs can be distributed however the Foundation and Tech & Ops Committee see fit
• We are currently in collaboration with the Business Operations team at Canton, please connect with @iliasc1 on Telegram for more specifics