The thin line between on-chain compliance and block censorship

Tara Annison
7 min readDec 15, 2020

“Blockseer’s new Bitcoin mining pool will be North America’s first bitcoin mining pool that will not only meet, but exceed the US Government’s Office of Foreign Assets Control (OFAC) compliance for BTC addresses, as well as providing the utmost level of transparency, auditability and corporate governance.” Oct 29th 2020: DMG Press Release

- — — — -

Since 1996 the United Nations Security Council (UNSC) has established 30 sanctions across a number of countries and regimes. These can come in the form of; economic, diplomatic, military, sport, environment and individual sanctions and they are adopted by many international bodies, such as the US Government’s Office of Foreign Asset Control (OFAC).

OFAC also holds a list of Specially Designated Nationals (SDNs) who are individuals and businesses found to have participated in illegal activities against the United States, such as money laundering, terrorism/terrorist financing, drug trafficking and human-rights violations. Typically these lists have focussed on activity within the traditional financial world, however growth in crypto use and price has seen some nefarious actors take advantage of this new financial system and use it to move the proceeds of their crimes. As such, the watchful eye of OFAC has turned to ensure that this activity is captured within their sanctions list updates and 2020 saw 4 OFAC announcements which included cryptocurrency addresses.

These were toward; Iranian nationals accused of cyber related actions against the US, Chinese nationals sanctioned under the Kingpin Act for manufacturing, selling, and distributing opioids in the US, North Korean actors accused of being part of the state sponsored Lazarus hacking outfit, and Russian actors said to be utilizing digital assets to channel funds in an effort to subvert the US elections.

Of these OFAC designations, 76.7% were for Bitcoin addresses, almost 7% were Ethereum addresses and just 6.8% were for privacy orientated protocols — Dash, ZCash and Monero.

Notably, whilst crypto addresses only featured in 4 of the 92 OFAC announcements for 2020, this is a marked increase from the 1 crypto related designation made in 2019, and it’s therefore very likely 2021 will see an increase in crypto-OFAC activity.

So where does Blockseer’s new mining pool come into all this?

Blockseer’s launch of their private beta mining pool includes the functionality that all mempool transactions (the waiting room for transactions which have not yet been included in a block: will be screened using their WalletScore blockchain analytics tool. Any which fail this check, whether due to interactions with illicit entities or through inclusion on the OFAC sanctions list, will not be included in their candidate blocks.

This sounds like a great way to ensure that sanctioned individuals aren’t using cryptocurrency to receive or move the proceeds of crime … but let’s unpack the implications of this a little…

Blockchain Analytics Tracing Shortfalls

There are many crypto-compliance services on the market which aim to provide actionable risk information through analytic exploration of blockchains. Given a transaction or an address, they’re able to provide information about the entities it has interacted with and where the funds have been received from or sent to. Whilst they may deploy different tracing approaches or methodologies, they’re all using the transparent nature of a blockchain coupled with their proprietary datasets to give risk based information. Blockseer’s WalletScore offers such functionality and therefore traces the history of a transaction or an address to check for illicit activity before deciding whether to include it within the candidate block. However consider the below subset of a blockchain graph;

Conducting a source of funds analysis on any of addresses C, D, E, F, G, H, I or J will result in some exposure to the OFAC address B. As such, any transactions involving these addresses would not be permitted in a Blockseer candidate block. However, the question which must be asked is whether an intentional transactional relationship can be drawn between each of these addresses and the OFAC labelled address B.

Clearly for C it can be, as it has directly received funds from the OFAC address and as all funds have moved directly from C to D one could quite reliably assume intentionality and a direct relationship there also. D then sends funds to an Exchange J and an unknown entity E, however tainting all transactions from Exchange J due to it’s proximity to OFAC address B would risk blocking many legitimate customer’s transactions from Exchange J simply because one customer has deposited OFAC related funds. In addition, the pattern of activity for E may suggest a user attempting to distribute their illicit-linked funds but could also be an unlabelled exchange. If the later, then addresses G, F, H and I may be tainted with exposure to the OFAC address B when in fact their exposure is better represented as to the unlabelled exchange E.

Therefore even looking at a small subset of a graph it’s very easy to see how the interconnectedness and potential lack of labels could result in an incorrect association with an OFAC related address and therefore exclusion from a block for any associated transactions.

This highlights the potential pitfalls of using such information to decide whether transactions are included in a miner’s candidate block as it could cause legitimate transactions by licit actors to be excluded from participation in the network. Furthermore, it emphasises the importance of good quality data being used and collected by blockchain analytics providers, with the risk that incorrect data could proliferate across the industry and impact decisions made for crypto-citizens.

Sanctions, but what next?

Whilst there is universal agreement about the criminal nature of sanctioned individuals, not all illicit categorisations are so agreed upon. Take for example the category of gambling which is illegal in the majority of US states and many countries around the world, but a legal and regulated activity in many others.

There is therefore a risk that different jurisdictions could seek to prohibit the inclusion of transactions from various categories or actors, and the result would be a patchwork of regulations across the world. This could result in miners selecting locations to open shop based not only on cheap electricity but also by the range of permitted transaction sources. This is likely to further centralized mining to regions with more permitting regulations and could see some jurisdictions as mining black holes where either no miners operate, or all do so anonymously — therefore denying potential tax income and blocking wider oversight from governments and regulators.

Technical Feasibility

As I detailed in my recent piece around the potential regulation of unhosted wallets, it is imperative that any new regulation or guidance around blockchain and crypto is both appropriate to manage the risk and technical feasible to implement and monitor.

In this case, whilst prohibiting the inclusion of transactions linked to sanctioned individuals (or other universally agreed nefarious actors) may successfully block their access to the crypto-ecosystem and therefore looks to mage the risk, one must question whether this would really be technically feasible to implement and monitor.

Firstly, whilst the location of many miners is known, there is a considerable portion of miners who mine anonymously.

As such whilst many miners choose to include identifiable information within the coinbase transaction, these unknown miners opt to collect their block reward and transaction fees without revealing their identity.

Trying to impose unilateral or jurisdictional regulations for block censorship would be very challenging to do as these unknown miners could create blocks with prohibited transactions and there would be little to no recourse to punish or stop them. In line with this, we may even see a greater swing towards unknown miners in the hashrate distribution since prohibited transactions would likely carry a higher transaction fee (to entice miners to include them into blocks) and therefore mining as an unknown miner to be able to include these transactions in your candidate block would be a more profitable venture.

Furthermore, any block censorship regulation would need to provide clarity upon whether miners could build on top of non-confirming blocks. Evidently the bitcoin blockchain’s history contains transactions linked to nefarious activity and actors, so would a miner need to have a certain number of ‘clean’ blocks before they can build a block on top? Or consider the scenario that the block at the tip includes gambling related transactions, would a US-confirming miner need to mine on the previous block instead or wait until the next ‘clean’ block?

Introducing such jurisdictionally sensitive regulation could create a highly volatile blockchain tip with many orphaned blocks, and would likely lead to a number of forks in which miners who are required to abide by block censorship rules create their own chains of fully compliant blocks. The Bitcoin blockchain would remain unscathed but the censored blockchains would probably meet the fate of the many forked and copycats of bitcoin who fail to attract community support and instead find their way to the back pages of CoinMarketCap. BlockTower’s chief information office Ari Paul foresees this and envisages a revolt by miners who could choose to orphan any blocks which censor transactions.

“There is a problem though — the 99% (or 51%) could choose to orphan any block with a blacklisted address, but this requires collusion.” (Twitter)

As such, any regulatory guidance around the transactions miners can include in their blocks should be treated with caution in order to avoid the unintended consequences of penalising legitimate crypto transactions or pushing regulated activity further from the bitcoin blockchain and instead into parallel censored chains.

(All views and information presented in this piece are the author’s own and do not necessarily reflect that of an employer, associated organisations or affiliations)