Charles Edwards, founder and CEO of Capriole Investments, has issued his starkest warning yet on quantum computing, arguing that Bitcoin must migrate to post-quantum signatures on an accelerated timeline or face existential risk later this decade. “We need to upgrade Bitcoin to be Quantum proof next year. 2026. Otherwise we are fucked,” Edwards wrote on X early Monday, escalating a series of posts in which he contends “Q-Day is this decade.”
Edwards’ timeline closely tracks a fresh line of research and industry messaging from Pierre-Luc Dallaire-Démers, founder of Pauli Group, a startup focused on quantum-resistant money. In an August research preprint and public thread, Dallaire-Démers and co-authors introduced graded ECDLP challenges on Bitcoin’s curve and, after translating logical circuits to physical costs across several error-corrected architectures, placed “cryptanalytically relevant” ECC-256 attacks in a “roughly 2027–2033” window—emphasizing wide error bars and sensitivity to hardware assumptions.
Pauli Group summarized the upshot plainly: “The first attack on 256-bit ECC will plausibly happen between 2027–2033.” The firm also provocatively stated via X: “PQC BTC will go to $1M+ by 2030. ECC BTC won’t.”
The core risk vector is well-established: once a Bitcoin address reveals its public key on-chain—by spending from it or by using legacy formats that expose the key outright—a sufficiently powerful quantum computer running Shor’s algorithm could, in principle, derive the private key quickly enough to steal funds.
In late 2024, material circulated in the NIST/PQ ecosystem sketched migrations away from vulnerable algorithms by roughly 2035—a horizon many security engineers view as realistic for broad IT systems, even if niche breakthroughs arrive sooner. The spread between the “thousands” versus “millions” of logical qubits camps reflects fast-evolving algorithmic optimizations, differing error-correction models, and varied assumptions about gate speeds and code distances.
At press time, BTC traded at $112,150.
