10 years into the blockchain revolution, how do we safeguard crypto assets today?
Soon, on 3 Jan 2019, it would be exactly a decade since the day when the genesis block of bitcoin was mined. Interestingly, this block included a header published at the same day in The Times: “The Times 03/Jan/2009 Chancellor on brink of second bailout for banks.”
Some have interpreted the choice to include this text as criticism on the instability of the current financial system. And while the global fractional-reserve banking system still dominates, it is astonishing to see how much progress was made since this first block. Cryptocurrencies are now officially a new asset class, with an aggregated market cap of hundreds of billions of dollars. Blockchain also seems to make its baby steps into the enterprise, with the first projects coming into production.
However, with all this progress, compromise of crypto assets takes place very frequently, with $1.1 billion stolen in cryptocurrency-related thefts in the first half of 2018 alone. While the industry matures rapidly with adoption from mainstream financial services players like Fidelity, which just launched an institutional platform for custody and trading of Bitcoin and Ethereum, it seems like securing crypto assets still has a long way to go.
In this short write up, we will examine why it is so hard to protect crypto assets, share how security practices have evolved and what the major and most advanced cryptocurrency service providers are doing to safeguard the assets they manage.
Four main factors that make custody of crypto assets so difficult
To put it simply, crypto assets of a certain account are represented by a balance stored in a ledger for a specific address. This address is hashed from a public key, and while it is public and everyone can deposit to it, only the controller of the private key can use the funds in this address and commit transactions.
1 — The key is the asset
With crypto assets, the key is the asset itself. This is very much different compared to traditional systems, where cryptographic keys are only pointers (tokens) to the asset, and NOT the asset itself. For example, when using an online banking app or a hardware token to move funds, the contained private keys are used for signing a transaction — however, the funds themselves are not stored in the mobile device or token, and thereby wouldn’t be harmed directly by the compromise of the private key.
The key being the asset has significant ramifications. Typically, the key itself is very small, e.g. only 32 bytes long in the case of Bitcoin. It’s hard to comprehend how much value can be stored in such a small and intangible series of bytes. Some Bitcoin addresses contain hundreds of millions of dollars– and only a 32-byte private key is used to claim ownership of this asset.
A fraudster compromising the key has complete control of the balance under a certain address; also, losing the keys for cryptocurrency is just as bad — as this British man found out, after throwing away a hard drive with private keys for over $80 million worth of Bitcoin.
2 — Use it once and lose it all
Actually, it’s much worse. A fraudster doesn’t even have to get hold of the private key in order to commit a fraudulent transaction. All it takes for committing such a transaction is a signature of the transaction with the private key. All it takes to empty an address on the ledger is a signature — whether it includes $40 or $100 million.
This fact cannot be undermined; while various methods for protecting cryptographic keys exist for a couple of decades, all of them — including the most robust hardware security modules (HSMs), focus mostly on preventing compromise of the key material. For many cryptography use cases, this is good enough, as the sheer damage caused by occasional key misuse is not catastrophic. However, this is not the case with crypto assets, where even a single use can mean “game over”: thus, even if keys are stored in an HSM, attackers are likely to target a much weaker link, i.e. the system which uses the HSM for signing transactions.
3 — No do-overs
From a fraudster’s point of view, one of the most lucrative aspects about cryptocurrencies is the irrevocability of transactions. Once a transaction was written to the blockchain, it cannot be undone. This is very different from the reality in the current financial system, where in many cases fraudulent deeds can be (at least partly) undone.
4 — Easy cash out
Easy cash out is literally a dream coming true for fraudsters. In committing traditional online financial fraud, cash out is by far the most difficult; that’s why the operators of banking trojans need to hire herds of money mules for concealing the money trail. Thus, the anonymity of crypto assets is heavily exploited by fraudsters, as it allows to steal and thereafter cash out very large amounts of money while remaining unnoticed.
It is not surprising why we see so much crypto assets stolen. Inherent weakness and vulnerabilities stem from the fact that secure key management is a very difficult task to carry out properly, combined with a massive motivation of fraudsters to exploit them due to the unforgiving nature of the blockchain combined with its anonymity. Nowadays, stealing cryptocurrencies is as close as it gets to the perfect crime.