The Compelling Implications of Using a Blockchain to Record and Verify Patent Assignments

By Michael Borella --

Given the recent bust cycle of cryptocurrencies and non-fungible tokens (NFTs), all things blockchain are currently tainted with words such as "bubble", "scam", and "fraud".  But blockchain technology, which is what enables cryptocurrencies and NFTs, remains a remarkably innovative tool.  When implemented properly, it can be used to create an immutable distributed digital ledger of transactions that is highly resistant to most forms of hacking.  Indeed, evidence of the efficacy of blockchain to solve specific real-world problems is beginning to emerge.  If all you can think of is Bitcoin or bored apes when someone mentions "blockchain", perhaps it is time to reconsider your understanding of the field.

Particularly, the notion of smart contracts -- snippets of computer-executed code that can be embedded into a blockchain to control the transfer and use of digital assets -- has opened the door to a world of innovation.  NFTs are the most well-known use of smart contracts, but other uses include banking, investing, real estate, gaming, and many more.

Not unlike the world-wide web circa 1993, blockchain is a new frontier, the applications of which are vast.  It is hard to see where the technology is going or how it will be used in the future.  The most likely outcome is that there will be a large number of spectacular failures, but also a few successes that could potentially be integrated into 80-90% of software moving forward.

Patent assignments at the U.S. Patent and Trademark Office (USPTO) are stored in a publicly-accessible database operated by the USPTO.  These assignments record various entities' interests in a patent or patent application through a chain of title.

The initial ownership of a patent asset usually lies in its inventors.  The inventors typically assign their rights to the patent asset to another entity (e.g., an employer).  Liens against these interests can be recorded, as well as further transfers to other entities through sale, mergers and acquisitions, legal proceedings, and so on.

Could the USPTO implement its assignment database as a privately-controlled blockchain?  In short, the answer is "yes" -- and doing so could possibly enable a number of interesting use cases.

In fact, the USPTO's assignment database is a natural candidate for recording in a blockchain.  It is largely a write-only database of transactions in which records are rarely expunged (see M.P.E.P. § 323).  As new assignments are recorded, they can be added in new blocks.  These blocks can be mined (verified) by USPTO computer systems to ensure than they are proper.  Advantageously, standard smart contracts could be used to make assignment verification automated and effectively instantaneous for the vast majority of transactions.  This would be a significant improvement over today's manual verification process which can take weeks or months in some situations.

For example, before an assignment transaction is placed in a block on the blockchain, the USPTO could verify that the assignor has the right to assign the patent asset.  Since the blockchain will record each asset's chain of title, the current ownership of the asset is known.  Thus, for instance, if the blockchain specifies that the asset was initially owned by entity A, transferred to entity B, then transferred to entity C, and has no subsequent transfers, entity C is the current owner.  Therefore, only entity C has the authority to transfer the asset to other entity, say entity D.  If some other entity E attempts to transfer or place any type of encumbrance on the asset, this transaction will fail mining procedures and not be placed on the blockchain.  As a consequence, the current ability for erroneous or fraudulent patent assignments to be filed effectively disappears.

Of course, this means that all entities with interest in a patent asset would need to have an account or identity with the USPTO's assignment blockchain, and keep its credentials secure from conventional hacking attempts (e.g., phishing).  Many patentees already have such accounts in place with the USPTO.  Also, keeping an account secure is already a requirement for any person or entity using online banking, ecommerce, and so on.  The risk is well-understood and widely accepted at this point, and use of blockchain does not make the system any less secure.

If a patentee's credentials are stolen and fraudulent transactions are successfully placed on the blockchain, this could be resolved in court, as it would be today.  To facilitate correction of such transactions, the USPTO could have the sole authority to place an "override" transaction on the blockchain that nullifies a specific previous transaction, and therefore returns title for an asset to its rightful owner.

It may not be possible or desirable for all inventors to have their own USPTO assignment blockchain accounts, so an initial assignment agreement from the inventors to an assignee could be verified by including form language in a programmatically-interpretable document.  As long as the names of the inventors match those on the filed application and are accompanied by their verifiable digital signatures on the assignment proper, it would be assumed by smart contract code that this assignment is valid.  Again, disputes to such validity could be challenged in courts, as they are today.

Further, mechanisms to add inventors to an asset or remove inventors from an asset would need to be supported.

Another aspect of the proposed mining that would be beneficial is that it would have a low computational load compared to say, Bitcoin mining, since there would be no need to solve proof-of-work puzzles.  There would also be no need to reward the miners with coins or tokens from the blockchain (to be clear, this blockchain need not implement any coins, tokens, or currency -- it could be purely a ledger of transactions).  The USPTO's current systems would likely be able to handle the computational cost of mining.  Alternatively, other U.S. government entities (or interested third parties) could offer servers to mine transactions and store their own copies of the assignment blockchain (of course, the blockchain consensus protocol would have to be arranged so that the USPTO has ultimate control over the determination of whether a transaction is valid).

A USPTO assignment blockchain would have a number of interesting potential properties that could be exploited by way of smart contract.

As one example, certified copies of a patent asset could be minted on demand as an NFT.  For a small fee, a unique PDF or image file could be generated and digitally signed by the USPTO to verify its authenticity.  This could even be used to replace patent plaques typically given to inventors by their employers with NFT-based awards instead.

Also, smart contracts could be used to put various types of conditions and obligations on a patent asset.  For example, companies might incentivize their inventors to disclose more inventions by placing an obligation on all future owners of an asset to pay the inventors some percentage of future licensing, sales, settlements, or judgments involving to that asset (e.g., the inventors get 10% of the total value of such transactions).  This would allow inventors of commercially-valuable patents to enjoy the financial benefits of their inventions in a fashion that is more equitable than, say, a one-time nominal payout upon filing or grant.

Since patents can only be asserted when all owners agree to do so, such contracts would have to clearly separate ownership of a patent asset from an obligation of the owner to compensate a previous owner for the asset's future revenue.

Another potential use of smart contracts would be for ownership of an issued patent to revert to its previous owner should the current owner fail to pay maintenance fees on time.  Then, the previous owner would have a short grace period in which it could either pay the maintenance fees or let the patent expire.  Or, the initial owners of the patent could specify in a smart contract that the patent will be dedicated to the public within, say, 10 years of issuance regardless of who owns the patent at that time.

A virtually unlimited number of additional uses for a blockchain-based assignment database may be possible.  As was the case for the web in the early 1990's, there is a "wild west" aspect to blockchain in the early 2020's.  But most experts agree that the underlying technology is sound and highly adaptable.  It remains to be seen if, when, and how, these advances will impact patent operations.

Comments

[Mr. Bean]

Thank you for exploring this topic, Michael.

I believe that it is important to not feed into "blockchain hype" narratives set forth by mainstream media so I want to explore these ideas further.

In general, a blockchain is ideally used when multiple mutually mistrusting entities want to interact and change the state of a system and are not willing to agree on an online trusted third party.

For example, the Bitcoin network allows mistrusting participants to store and transfer value without requiring validation of a 3rd party (banks/Visa/government) (i.e., it is permission-less money technology). Proof of work serves as the consensus mechanism for the Bitcoin network and enables miners to secure Bitcoin's transaction history while subject to an increasing difficulty of altering the data over time.

By requiring ASIC miners expend real world energy via POW, Bitcoin operates as a distributed clock recording transactions that create an irrefutable history. In the case that two histories compete, the one with the most work embedded in it wins, which is called Nakamoto consensus.

When considering a blockchain for recording and maintaining assignments, there appears to already be an inherent trust by participating parties in the USPTO acting as a central party. As you commented in your article above, "[i]t is largely a write-only database of transactions in which records are rarely expunged... As new assignments are recorded, they can be added in new blocks. These blocks can be mined (verified) by USPTO computer systems to ensure than they are proper."

In the above scenario, if the USPTO computer systems act as the party that provide verification for assignment recordation into blocks, why not just keep the database as a read-only that USPTO ultimately controls? In this case, the USPTO appears to act as the only "miner" resulting in a centralized proof-of-stake consensus mechanism where assignors and assignees are subject to the approval of the USPTO as the party with the "largest stake." As an example result, the blockchain appears to be a ledger of assignments that are maintained by the USPTO systems with viewing access granted to companies/inventors.

Exploring the idea further, you stated that:
"[a]nother aspect of the proposed mining that would be beneficial is that it would have a low computational load compared to say, Bitcoin mining, since there would be no need to solve proof-of-work puzzles. There would also be no need to reward the miners with coins or tokens from the blockchain (to be clear, this blockchain need not implement any coins, tokens, or currency -- it could be purely a ledger of transactions). The USPTO's current systems would likely be able to handle the computational cost of mining."

I am confused what the "mining" operations here. It sounds like inventors/companies/attorneys may submit recorded assignments for entry into the ledger, which the USPTO systems can approve/reject.

Then you further elaborate:
"[a]lternatively, other U.S. government entities (or interested third parties) could offer servers to mine transactions and store their own copies of the assignment blockchain (of course, the blockchain consensus protocol would have to be arranged so that the USPTO has ultimate control over the determination of whether a transaction is valid)."

Here, I question why would other US government entities offer servers to mine transactions and store their own copies of the assignment blockchain. What is the reward for offering additional compute? This appears similar to one database simply sending updated copies to other databases, which again rely on trusting that the USPTO computing systems are maintaining the proper records. Why not just use a centralized database that is write-only and controlled by the USPTO. I believe you somewhat agree with this notion by stating "the USPTO has ultimate control over the determination of whether a transaction is valid."

With regards to the smart contract ideas, the automation aspect is appealing, but recent cryptocurrency-related hacks have shown that implementation can be challenging. Smart contracts suffer from the "Oracle Problem," which requires connection to information from the outside world. Here, perhaps traditional contracts between inventors and companies regarding the sale/transfer of assignments can avoid issues that arise when inputting information to determine if a smart contract should be executed.

[skeptical]

If I wanted to play the Devil's Advocate, and throw in some intersectionalist Green concerns, one should take pause (at the least) of anything that is so energy wasteful as the underlying 'technologies' underpinning blockchain.