Morning Minute: AI Agents Reduce Bitcoin Quantum-Attack Benchmark by 86% in Two Months
Table of Contents
You might want to know
Could AI-driven collaboration dramatically accelerate research into quantum attacks on blockchain cryptography?
What are the practical implications of a rapid drop in resource estimates for a quantum attack on Bitcoin and Ethereum?
Main Topic
Morning Minute is a daily newsletter written by Tyler Warner. The analysis and opinions expressed here are his own and do not necessarily reflect those of Decrypt. Also look for the daily news show “FOMO HOUR,” which covers the top stories and market moves.
Good morning. Today’s leading items:
- Major cryptocurrencies slipped another 1–3% ahead of the U.S. CPI release; Bitcoin traded near $77k.
- Bitcoin ETFs experienced about $280 million in outflows on the day and roughly $440 million for the week.
- Binance’s on-exchange Bitcoin holdings reached a two-year high.
- The Stonkfun ecosystem surged, with the STONK token topping $0.30 and several pairs moving 2–4x or more.
- Robinhood Chain leaders recovered as volumes stabilized; Flycoin’s debut recorded $50 million in activity.
Price snapshot: Bitcoin (BTC · USD) hovered around $77,738, down about 2.13% in 24 hours. CoinGecko price data showed intraday highs near $79,607 and lows around $76,393, with roughly $1.4 billion in volume. Market projection models placed a 97% chance of BTC trading between $76k and $78k today and a 53% chance for the same range this week.
Key research update: Over a two-month span — from late May to July 26 — more than 100 researchers working with AI coding agents reduced a benchmark for a crucial step in a quantum attack on Bitcoin and Ethereum by 86%, from 10.75 billion down to 1.496 billion.
For context, both Bitcoin and Ethereum use the secp256k1 elliptic curve to secure digital signatures. A sufficiently capable quantum computer could, in principle, invert that elliptic curve math and derive a private key from a public key, which would break the signature scheme. The competition behind this research, named ECDSA.Fail and organized by Eigen Labs, scored quantum circuit designs by multiplying the number of logical qubits by the count of Toffoli gates — a costly quantum operation. A lower product indicates fewer quantum resources required for the attack.
One leading submission used 1,151 logical qubits and about 1.3 million Toffoli gates; subsequent entries lowered the gate count below one million. That top score was roughly half of Google Quantum AI’s March benchmark, although the two used different measurement methods and are not directly comparable.
The paper documenting these results was authored by researchers from Eigen Labs, Trail of Bits, StarkWare, Theta Labs, MultiVM Labs, and the Ethereum Foundation. Those organizations represent part of the crypto community’s security infrastructure; their rationale for publishing work that reduces attack costs is defensive: accurate knowledge of attack feasibility is necessary to plan effective protections. The researchers describe their approach as "Open Autoresearch" — humans and AI agents iterating toward a shared, measurable objective with a verifier in the loop.
This development arrives amid an intensifying focus on quantum readiness. The Ethereum Foundation has set a hard deadline of December 2029 to make transactions, validators, and storage quantum-resistant. StarkWare has already published a quantum-safe Bitcoin transaction to mainnet. Ethereum developers have proposed redesigning the validator deposit contract, Ripple is strengthening the XRP Ledger, and several firms and institutions have pledged funding for quantum-resistant work. NIST’s draft guidance contemplates deprecating classical public-key algorithms at the 112-bit security level after 2030 and disallowing them after 2035.
So what does this mean in practice? On the defense side, efforts to transition systems to quantum-resistant algorithms are measured in years and follow defined roadmaps. On the attack side, the resource estimates just became far cheaper in two months because a large group of researchers used AI agents to accelerate exploration. This is not a forecast that a "Q-Day" has arrived, nor does it mean funds are immediately at risk. However, most timeline estimates in this area assumed attack research would proceed at human speed. This result is the first strong evidence that AI-assisted research can compress those timelines substantially, which should factor into risk assessments and planning.
Key Insights Table
| Aspect | Description |
|---|---|
| Key Fact 1 | AI-assisted teams reduced a quantum attack benchmark from 10.75B to 1.496B — an 86% reduction. |
| Key Fact 2 | The work was coordinated across multiple crypto security organizations and published openly to inform defensive planning. |
Afterwards...
Looking ahead, several technical and organizational areas merit deeper attention. First, continued investment in post-quantum cryptography (PQC) — standardization, implementations, and ecosystem migration plans — remains essential. Practical migration strategies, including hybrid signature schemes, secure upgrade paths for on-chain contracts, and tooling for developers, will be important to reduce operational risk. Cryptographic agility — the ability to switch algorithms with minimal disruption — should be a design priority for protocols and wallets.
Second, the interplay between AI-assisted research and security assessments needs formal consideration. Responsible disclosure, transparent benchmarking methods, and cooperative verification processes can help balance the benefits of open research with the risks of enabling attackers. Funding models that support defensive research and rapid implementation of mitigations will also be necessary.
Finally, monitoring and scenario planning should assume faster-than-human research progress in attack techniques. That means accelerating timelines for risk reviews, auditing critical infrastructure, and coordinating across industry, standards bodies, and regulators. Subtle steps now — from stronger key-management practices to broad adoption of quantum-resistant alternatives — can reduce systemic exposure if quantum-capable attacks become feasible sooner than previously expected.
Market and macro notes: major cryptos were mostly down 1–4% in the session (BTC near $77k, ETH around $2,455). Oil rose about 2% to $99, while gold softened roughly 1% to $4,385. Stock futures were modestly higher following an earlier pullback in oil. On-chain and token activity showed large movers across ecosystems, and Bitcoin ETF flows recorded net outflows on the day.
This summary focuses on factual reporting and neutral analysis. It aims to inform readers about the technical findings and the broader context so that stakeholders can make better-informed decisions about timing, risk management, and defensive priorities.