Lost Your Keys, Lost Your Coins: Zcash Founder Wants Bitcoin to Stop Pretending 21M Is Real 🗝️

Eli Ben-Sasson, founder of Zcash and CEO of StarkWare, the company behind Ethereum Layer 2 scaling solution Starknet, publicly argued on July 7, 2026 that Bitcoin's 21 million supply cap "doesn't make sense," proposing instead that the network adopt a hard ceiling on its annual issuance rate. The proposal would replace Bitcoin's fixed total-coin ceiling with a fixed inflation rate ceiling, with Ben-Sasson naming 4% per year as a specific figure and describing it as "a reasonable upper bound on human population expansion."

Ben-Sasson's core argument centers on key loss. Because private keys are permanently lost over time, the coins attached to those keys remain on the ledger but fall out of practical circulation, making the usable supply unknowable and trending downward. "Capping the supply of Bitcoin at 21M doesn't make sense. Beacuse over time, keys will be lost. In fact, as time goes to infinity, all keys will be lost," he wrote on X on July 7, 2026, adding that he strongly supports "a clear monetary policy with an absolute upper bound on the # of Bitcoins in the future."

The proposal marks a conceptual shift from capping the stock of coins to capping the annual flow of new issuance, a distinction that carries structural implications for $BTC holders who have priced Bitcoin's fixed scarcity into their position. By transitioning to an issuance-rate ceiling rather than a unit ceiling, Bitcoin's monetary policy would be measured against yearly emission rather than a terminal coin count.

Ben-Sasson co-founded Zcash and currently serves as CEO of StarkWare, which develops the Starknet rollup on $ETH. His comments frame the 21 million figure as a nominal rather than effective cap, since lost coins reduce the spendable supply without removing units from the chain. The 4% figure he cited aligns roughly with long-run projections of global population growth rather than with any existing $BTC issuance schedule, which currently cuts in half approximately every four years through programmed halvings.

The remarks were published as commentary on X and have not been tied to any formal Bitcoin Improvement Proposal or coordinated network effort, and no Bitcoin Core contributors publicly endorsed the framing at the time of publication.