A covenant enforced through signing
A Bitcoin covenant is a rule that limits how particular coins can be spent. Cointelegraph Research describes a hypothetical approach using functional encryption (FE): instead of changing how Bitcoin nodes validate signatures, the design would restrict which transactions a key can sign.
Under the proposal, the key would produce signatures only for transactions in a predefined set of valid spends. Nodes would then check those signatures with OP_CHECKSIG as usual. Cointelegraph Research says this would avoid the need for a soft fork; it describes a proposal, not a feature deployed on Bitcoin.
How the proposed key restriction works
The mechanism is more involved than simply encrypting a private key and asking it to sign. Cointelegraph Research says the function would use an encrypted private key, k, but would not sign directly with k. Instead, it would derive a new key pair by tweaking k with a hash of the covenant’s spending policy. The policy therefore becomes part of the key used for signing.
That setup introduces a trust requirement. Cointelegraph Research says the setup authority must destroy both k and the master secret key, msk. If it keeps them, it could sign spends that violate the covenant, defeating the restriction.
Why the cryptography remains impractical
The central obstacle is the cryptographic construction needed to make the signing function enforce the policy. Cointelegraph Research says the required primitive is at least as hard to build as indistinguishability obfuscation, and that implementations of the principle remain at toy scale.
The article illustrates the scale gap with work by Carmer, Malozemoff and Raykova (2017). It reports that obfuscating a 12-bit pseudorandom function—a toy circuit of about 62,000 gates with a 64-bit key—took 3.7 hours. The resulting program occupied 120 GB, and one evaluation took 67 minutes with 128 GB of RAM. The article says this circuit was far smaller than ECDSA signing on 256-bit numbers.
There is also a limitation in the kinds of functions available from practical FE schemes. Cointelegraph Research says these schemes cover narrow classes, such as inner products, but not the ECDSA signing needed for the proposed Bitcoin covenants. These limits are why the mechanism remains hypothetical rather than a practical spending feature.
A proposed alternative
Cointelegraph Research points to a later discussion of “Un-FE’d Covenants,” which Rubin proposed as an approximation of covenant behavior without relying on impractical FE schemes.