Why Institutional Custody Breaks on Shielded Chains
summary
Institutional custody fails on shielded chains because threshold signatures don’t apply to ZK proofs, and the proposed fix is to run the prover inside an MPC. The claim: co-SNARKs make proofs thresholdizable so a proof verifies on-chain unchanged. TACEO has built this stack for 2.5 years, a network live two months doing 100,000 tx/day, with co-Circom and Noir tooling and low overhead; the market stakes are large (Coinbase ~$300B, Fireblocks $10 trillion volume), and the system works in production.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 0:12What TACEO Is Building
- 0:30A Problem Hidden in Plain Sight
- 0:52The Scale of Custody Today
- 1:25Fireblocks and $10 Trillion in Volume
- 1:52The Privacy Renaissance
- 2:32Why These Models Assume Self Custody
- 2:57Why Single Key Custody Fails Institutions
- 3:24How State of the Art Custody Splits Keys
- 4:01Threshold Signatures and Multi-Party Computation
- 4:35Why ECDSA Is Commonly Thresholdized
- 5:08Where Things Break in the ZK World
- 5:43Collaborative SNARKs From Academia
- 6:17How Co-SNARKs Thresholdize Proving
- 6:39How Far TACEO Has Come
- 7:11Tooling for Co-Circom and Noir
- 7:30Why the Performance Overhead Is Low
- 7:51A Call to Custodians
- 8:26Closing
related talks
01h 00mThe Path to Shielded Programmability
24mFrom Custody to Capital Markets Infrastructure
54mHalo2 circuit changes for Zcash Shielded Asset and Atomic Swaps
05mCompany Pitches
32mSoK: Private Programmability on Blockchain Systems
35mChains of trust in Confidential Computing
01h 01mThe Future of Zcash Panel
52mBinding Blockchains Together with Accountability