What counts as "broken":

A publicly demonstrated or credibly reported derivation of a private key from a public key for a standard secp256k1 key (256-bit ECC), on Bitcoin mainnet or in an equivalent laboratory setting. The attack must be computational — social engineering, wallet implementation bugs, or theft of key material do not count.

Answers:

• Quantum computational — Shor's algorithm or a quantum-native discrete logarithm attack. Quantum hardware is essential to the attack. Classical pre-processing is fine as long as the quantum step is doing the core work. If AI is used only to optimize the quantum implementation (error correction tuning, pulse scheduling, circuit depth), this still resolves here.

• AI / Classical breakthrough — The full attack runs on classical hardware (CPUs, GPUs, TPUs). Includes AI/ML-discovered algorithmic shortcuts, novel mathematical attacks on ECDLP, or large-scale side-channel methods.

• AI-accelerated quantum hybrid — AI discovers a mathematical reduction that changes the algorithmic complexity of the problem (e.g., reduces the required qubit count by an order of magnitude or more), but quantum hardware remains essential for the final key recovery. The distinction from "Quantum computational": the AI contribution must affect the algorithm, not just the implementation.

• Not broken by 2030 — No such attack is demonstrated before Jan 1, 2030. Includes the scenario where Bitcoin migrates to post-quantum cryptography (BIP-360, SPHINCS+, or similar) before any break occurs.

Edge cases:

• If multiple methods succeed close together, the first publicly verified attack determines resolution.

• Breaking SHA-256 (PoW/mining) does NOT count — this is specifically about ECDSA/Schnorr key recovery on secp256k1.

• If a nation-state break is only revealed later, resolution is based on when credible public evidence emerges.

Context (April 2026):

• Google Quantum AI estimated <500k physical qubits needed to break 256-bit ECC (down ~20x from prior estimates); Oratomic/Caltech estimated ~10k–26k with neutral-atom architectures. Current largest processors: ~1,000–1,500 physical qubits.

• ~6.9M BTC (~$440B) sit in addresses with exposed public keys. Bitcoin has not begun a post-quantum migration; last soft fork was Taproot (Nov 2021).

• Justin Drake (EF) estimated ≥10% chance of quantum key recovery by 2032.