Provenance and Intellectual Property (IP) Rights in the Blockchain Ecosystem
summary
Panel argues that blockchain can record who contributed data, compute and model recipes and enable revenue splits (examples include proposals like 5% GPU share, 10% data share) but fine-grained attribution is costly and unreliable because models are stochastic and per-output tracing runs 5–10× attribution cost. Work underway includes a prototype blockchain provenance layer for AI lifecycle and marketplace ideas (on-chain licensing, DeFi secondary markets); existing precedents cited are NFTs and Bored Ape IP transfer showing community-driven monetization.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:05AI sovereignty and blockchain provenance
- 3:52AI-generated assets as IP
- 7:06Attribution challenges in AI models
- 10:18Blockchain as global trademark office
- 13:29Decentralized vs centralized AI compute
- 17:23Permissionless IP rights management
- 20:14Key progress in AI and blockchain IP
- 21:59Future of creator revenue
related talks
21mNon fungible tokens and IP issues
10mWhy the Future of AI Depends on Blockchain
30mIP Expansion through Blockchain Infrastructure
49mBlockchains for Artificial Intelligence
31mA New Global IP Infrastructure in the Age of Generative AI: Git for IP
20mIP - Biggest Problem in the Age of AI
24mNFTs and Intellectual Property Rights: real life use cases
57mRewiring the Internet for Ownership with Big Data and Blockchains