Half-Baked NFT Licensing Protocol Idea

INTRO: I AM LAZY BUT CREATIVE

A lot of lawyers are now talking about NFT IP & licensing issues. I’m not much of an IP lawyer or NFT collector, but I did write a previous thing about NFTs here, and I worked with Rarible for a while and helped them draft this idea of a standard NFT bearer-style license. Since a lot of people are now talking about this stuff and I had sort of a half-baked idea that I can’t afford to bake the rest of the way, I thought I’d share it quickly and, who knows, maybe someone else will pick it up or be inspired by it since I can’t really follow through.

First I should explain how I see NFTs. Basically I think they’re either dumb or genius. They’re dumb if they just rely on “rough social consensus” for their value. They’re genius if they marry legal engineering and cultural engineering to increase social scalability in the realm of art.

To be a bit more specific, my basic idea about NFTs is roughly along the following lines:

• No one currently thinks IP rights matter because we’re in an NFT bull market where valuations levitate due to hype and “social consensus”, but this won’t last forever.

• But when the euphoria dies down and as time goes on and market matures, everyone—artists, collectors, platforms—will realize that the IP issues for NFTs matter a lot—not only due to issues like lawsuits over copyright and takedown requests, but also for preservation of the NFTs’ market value.

• Ultimately, NFTs are trending toward composability—for example, the ability to combine a base LOOT NFT with a base CHARACTER NFT to create a character with items—and I believe the IP lawyers’ like to call these “derivative works”—which typically require at least a license to create, so IP rights are important for this reason as well.

• Just like everything blockchain-related, ultimately NFT/art combos are only useful to be on the blockchain if putting them on the blockchain increases social scalability by providing stronger security (trust-minimization) guarantees than are available through centralized distribution/sale/collection/holding methods for art.

Based on this philosophy and outlook, I propose an NFT licensing protocol along the following lines:

TLDR: 

• fork existing NFT standards to add a "license" metadata field which stores hash of license

• encourage license standardization with limited variables so that there are a limited # of hashes whose meaning can be recognized on-chain

• a set of wrapper contracts is deployed; these contracts are compatible with the standard license

• the wrapper contracts "know" which kind of standard license each standard NFT is associate with (i.e., how the standard license was parametrized), just by "checking" the license field of the NFT  

• if IP license in "license" field of NFT allows for the kind of sublicense represented by the wrapper contract, wrapper contract will accept that NFT for wrapping; in return, the wrapper contract gives the NFT owner a new NFT which means that person still owns the NFT, but the NFT is encumbered by a license ("authNFT") and gives the sublicensee a sublicense token which means that sublicensee has certain rights to use the associated IP/art ("licNFT") 

• if IP license in "license" field of NFT does not allow for the kind of sublicense represented by the relevant wrapper contract, the wrapper contract will not accept deposit of that NFT and will not produce authNFT and licNFT  

• the protocol should have a governance token so license standard can evolve and grow more parametrizable over time 

• the standard license MUST have the following terms (i.e., these terms must be in standard license for the protocol to function as designed): 

  • holding the NFT (or one of the possible licNFTs) means holding the license on a "bearer basis"

• I suggest the standard license SHOULD also have the following license terms (i.e., these are not essential to use of the above protocol scheme but is my personal view on a "good standard" to drive the market toward): 

  • strong constraints on further discretionary artist use of that exact IP after initial sale of NFT 

  • if IP is part of a "family/brand of NFTs," the "family/brand IP" is governed by all NFT holders for that family/brand 

  • collector subrogated to artist rights to enforce DMCA takedown requests etc.

Longer Explanation + Examples

• Target Functionality 

  • licensing protocol for the IP embedded in art, pfp & badge NFTs 

    • I.e., licensing protocol designed for use with NFTs with metadata which ties the token to off-chain visual IP (art/design/badge) stored off-chain (assume IPFS)

• Design Goals

  • encourage market standardization by defining “the right NFT license”, providing  limited adjustable parameters and developing awesome infra (i.e., an associated protocol) which makes market converge toward adopting that license 

  • enforcing a specific philosophy regarding how the target NFT market should work and what makes target NFTs valuable:

    • Minimize trust, which dictates the following design choices: 

      • “bearer IP rights model” where merely owning an NFT is strong legal evidence that the holder has certain IP rights in the art/pfp/badge

      • maximize legal and on-chain guarantees of rarity

        • this entails constraining artist discretion after sale of the NFT 

          • At a minimum, the artist can’t make arbitrarily make more NFTs

          • Consider whether artist should be allowed to make other uses of the IP or whether the artist should be severely constrained in future uses of the IP–for example, because the artist printing a million tshirts with the same art could dilute the value of the NFT 

      • minimize human judgment/interpretation/discretion 

        • This dictates, for example, that machine systems can “know” what license is involved merely by looking at on-chain data, despite that the license is a human-readable legal agreement stored on IPFS–this achievable with standard agreement variants that therefore have standard hashes  

    • maximize interoperability

    • maximize composability

    • maximize extensibility

    • Ensure sustainability /evolvability of protocol

      • Provide for a source of funds for ongoing work/improvement to the protocol

      • Provide for governance 

• Implementation 

  • Possible License Terms - STRONG Collector Protection Version 

    • License similar to exclusive Rarible license

    • License = exclusive license to the NFT holder to do whatever they want with the specific art or other IP the NFT is meant to represent (including sublicense it), subject to possible governance override (if the NFT is part of a family and that family has its own governance) 

    • This is a bearer license–Lawfully holding & owning the NFT = holding & owning the license to the IP (“lawfully” is required so that if NFT was stolen, the thief does not have the license rights) 

    • Artist cannot do anything else with the IP ever again without the consent of the holder AND/OR community of holders of that family of NFTs (implement some governance standard–e.g. majority)

      • Obviously, lots of ways to skin the cat here. . .and not all NFTs will be part of a family…not all families will have governance…. 

    • Holder is subrogated to artist’s rights to enforce IP rights - - - e.g., holder can make DMCA takedown requests

    • If the NFT is part of a family with governance, then the family ‘DAO” also is subrogated to such IP enforcement rights 

    • Each owner can do anything with the design specific to that holder’s specific NFT, but can be overridden by NFT family governance (e.g., because what they are doing “dilutes the brand” and thus adversely affects the value of others’ NFTs)

    • Need to include some rules around cross-chain and (contentious) hardforks 

• On-chain features

  • Forks of ERC 721 and ERC 1155 (referred to in this spec as “FERC-721,” “FERC-1155,” and collectively “FERCs”).  FERCs have one key difference: they provide a new “license” field for the metadata, which is filled with a hash of the selected license (potentially with a few adjustable parameters that the artist can choose in the minting process–thus different hashes for each variant)

    • Query whether the standard license should be subject to “protocol governance” and thus can be updated over time (thus, by using these forks of ERC 721 and ERC 1155, the artist and all buyers agree to be part of this governance and to be subject to such license updates) 

  • Wrapper contracts that “know” which license each FERC is subject to and symbolize the granting of a permissible sublicense 

  • Wrapper contracts would come in standard “flavors” and could even automatically handle royalties–for example, if the sublicensee fails to make a regular payment then the wrapper could automatically be unwound (automatic enforcement of holder’s right to terminate the sublicense) 

  • Wrapper contracts could also be directly subject to family NFT governance, furthering trust minimization 

• Meta–Licensing Protocol Value Accrual and Governance 

  • Charge fees to users of this licensing protocol

  • Have governance token for this licensing protocol 

  • Liquidity mining of governance token for users of the protocol 

  • Governance can update the “standard license” (FERC standard), any sublicense (the wrappers) or add new standard licenses (FERC standards) or sublicenses (wrapper contracts)

• Example #1 - sublicense to display art in museum: 

  • Holder buys FERC, which has the standard license parameterized to allow the holder to sublicense the right to display a digital copy of the relevant art on a screen in a museum 

  • Holder decides that in fact it wishes to exercise the sublicense right by granting MoMa the right to display the FERC art on a screen in MoMa

  • Holder stakes FERC in “standard display rights” wrapper contract: 

    • Holder receives dispFERC, which is redeemable for Holder’s specific FERC at any time Holder wants to un-stake (thus, dispFERC should have the same market value/price as the staked FERC) 

    • MoMa receives subdispFERC, which is tied to a standard sublicense agreement (as defined by the wrapper contract). MoMa can use its control of subdispFERC as evidence that it has the legal right to display the art 

  • MoMa displays the FERC art next to a picture of Hitler, devaluing the FERC and angering the Holder 

    • Holder redeeems dispFERC for FERC 

    • Under the logic embedded in the wrapper contract, MoMa’s subsdispFERC is automatically burned

    • Under the terms of the license and sublicense, and pursuant to the functionality embedded in the wrapper contract, MoMa has lost the right to display the FERC art and can be legally compelled to stop displaying it 

• Example #2 - attempt to sublicense, but FERC license does not allow that kind of sublicense

  • Same hypo, but this time the relevant FERC’s license (as represented by the hash in the license field of the FERC) is not parameterized to allow sublicensing of display rights 

  • Holder tries to stake FERC in “standard display rights” wrapper contract 

  • Contract is not set to accept a FERC parametrized not to allow this sublicense, and rejects the staking 

  • Transaction reverts, Holder still has FERC, MoMa cannot obtain display rights through a sublicense but must go back to the original artist and negotiate a deal 

CONCLUSION

I have literally no time to further pursue this idea, but if you think it’s cool and do something with it or something inspired by it I’d always appreciate some credit and updates!