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.