TL;DR: Coinbase says it was not affected by the reported Binance hot-wallet incident and that more than 99% of customer assets are kept in cold storage. The exchange outlined controls around signing, withdrawal limits and suspicious activity, offering a useful test of how custodians should explain security practices.
A major attack on one exchange inevitably prompts customers elsewhere to ask the same question: where are my assets actually held, and what stops them being moved without permission? Coinbase addressed those concerns in a November 4 security explanation, published against the backdrop of the reported Binance incident.
In its official post , Coinbase said it had not been affected by the breach and that more than 99% of customer assets were held in cold storage. It also described distributed signing keys, withdrawal limits, anomaly detection and emergency controls capable of stopping transaction signing.
What cold storage protects, and what it does not
Cold storage refers broadly to arrangements designed to keep signing authority away from routine internet-connected systems. It can reduce the exposure of a custodian’s largest reserves if a hot wallet or application layer is compromised. Hot wallets remain useful for normal withdrawals, so the balance between operational liquidity and offline custody matters.
The reported percentage is Coinbase’s stated current practice; it should not be interpreted as a universal guarantee that every customer balance is insulated from every type of problem. Account takeover, internal controls, fraud, software defects and legal restrictions are separate categories of risk. A security system is strongest when the organisation explains the entire chain of access rather than relying on one impressive percentage.
Similarly, distributed keys can mean that no single compromised system can authorise a large movement, but only if the policy around those keys is well implemented. Limits and anomaly detection add further layers, potentially buying time when a transaction looks unusual.
Why the industry is revisiting wallet architecture
Custodians face an awkward operational challenge. Users expect transfers to work continuously, but allowing automated movement from the internet-connected part of a system increases the consequences of a compromise. A good model therefore combines preventative controls with monitoring and rehearsed incident procedures.
The disclosure comes as industry participants investigate the reported multi-billion-dollar Binance exploit and as stablecoin issuers describe the actions available to them once assets have left a platform. These are connected but distinct defences: custody aims to stop unauthorised signing, while downstream tracing and freezes aim to limit damage after funds move.
The importance of resilience is also visible in decentralised finance. As NewsBTC reported when Aave restricted new tBTC supply and borrowing , operational controls can reduce further exposure during an incident. They are most useful when their scope is explicit and their effect is independently checked.
What exchange users should ask
Useful questions include whether balances are fully backed, what proportion is held in internet-connected wallets, how withdrawals are approved, whether controls are independently assessed, and how the company communicates if access is interrupted. A customer should also understand the difference between assets held by a platform and tokens in a wallet whose private keys they personally control.
Coinbase’s statement offers concrete details rather than simply calling itself secure. Those details still need to be assessed over time against audits, incident disclosures and actual operational performance. The wider lesson is that custody design belongs at the centre of an exchange’s risk story, not in small print below a trading app.
This article was written by the News Desk and edited by Samuel Rae.
