Privacy is evolving before our eyes
Reflecting on the DC Privacy Summit
We’re back! Thank you so much for your patience while we got our first (hopefully of many!) in-person event off the ground. And thanks to all who attended the DC Privacy Summit! If you missed it this year, don’t worry. It’s coming back in 2025.
Reflections from the DC Privacy Summit
In biology, selective pressure is a force, often a characteristic of the environment, that drives evolution. Nature selects individuals with traits that keep them from succumbing to disease, predation, starvation, and the like.
In the same way, the actions of governments have helped drive the evolution of decentralized systems. First, agencies learned how to follow the money on blockchains, inspiring the creation of novel privacy technologies. Then developers of such novel privacy technologies were branded criminals. That has catalyzed the creation of an even newer class of sophisticated technical tools aimed at keeping bad actors out of decentralized systems.
It is this technological evolution that we placed under the spotlight late last month when we held the inaugural DC Privacy Summit conference. The question remains whether policymakers will evolve in response.
The Parable of Clearville
If you’ve followed this newsletter for any length of time, you probably know that we are obsessed with the case of Tornado Cash, the Ethereum-based privacy tool that uses state-of-the-art cryptography to anonymize cryptocurrency users. What got us so interested? It was the way governments reacted when they learned that the North Korean regime was allegedly using Tornado Cash to launder millions of stolen crypto-dollars. For the first time ever, the US imposed sanctions on a set of smart contracts. Law enforcement officials portrayed the developers of the open-source privacy tool as criminals—in part because the developers allegedly knew the North Koreans were using the tool and didn’t stop it.
In fact, the tool is a set of smart contracts that the developers have no means to shut down or modify. People are still using it despite the fact that one developer, Alexey Pertsev is serving time in a Dutch prison for allegedly facilitating money laundering. Another, Roman Storm, faces more than four decades behind bars if he’s convicted in a US trial next year on charges of money laundering, violating US sanctions law, and operating an unlicensed money transmitter.
While casting Tornado Cash’s developers as criminal money launderers, law enforcement officials and some policymakers have also implied that Tornado Cash itself is a money laundering tool. In the view of most proponents of cryptocurrency and blockchain networks, that’s like saying a steak knife is a tool for stabbing people. It’s true that it’s possible to pick up a steak knife and go commit a stabbing, just as a shadowy cyberthief might look to Tornado Cash to cover their tracks. But that’s not why the tool was invented.
Why was it invented? Because what happens on Ethereum is visible to everyone—almost like a town composed completely of transparent glass houses, Peter Van Valkenburgh, cryptocurrency policy advocacy group Coin Center’s director of research, argued in his opening keynote at the DC Privacy Summit.
Well, he did more than argue that. He told a story in the style of Dr. Seuss: The Parable of Clearville, the Land of Glass Houses. “The townsfolk of Clearville were very polite. They knew they were watched from morning through night,” he began. “They also were a bit dull. An entire town in a creative lull.” One day a stranger arrived in an opaque truck. His name was Shade, and he was hawking a compelling new product: “Curtains, blinds, even frosted glass. And sold them he did, right from his truck. He sold the lot to the last.”
“You see, the Clearvillians were pretty darn sick of always looking over their shoulders, even in the relative privacy of their glass people holders. A few said ‘Well, I’ve got nothing to hide. But even those folks admitted that a curtain or two in the bathroom was good for their pride.”
The police chief, Constable Eye, was bewildered. “It was once the case that Eye would simply walk by and, well… spy… to be sure that nothing so sly was going on in the houses of Clearville.” Now the shades blocked his view, so he started banging on doors. But “you can’t just barge into every house that’s suspicious,” Clearville’s mayor told Eye. “You’ve got to have evidence your hunch is propitious.”
Eye had a new plan: “We’ll simply go raid the truck of that despicable Shade.”
But “I just gave them some shade,” pleaded Shade, in response to the raid. “Just what they might do, even inside of my view, was none of my business or trade.”
“So you saw what they did, and it’s that which you hid,” Eye replied. “What a very fine tale. Now we’ll take you to jail, and of this troublesome truck we’ll be rid.”
New systems challenge old laws
Like the new materials that Shade brought to Clearville, Tornado Cash and systems like it render traditional ways of managing risk and keeping regular people safe less effective. That argument is core to a compelling policy proposal introduced earlier this year by veteran crypto lawyers Michael Mosier and Rebecca Rettig. Mosier was a federal financial crimes regulator and crypto startup general counsel and now runs a boutique law firm called Arktouros. Rettig is the chief legal and policy officer at Polygon Labs. Together with Katja Gilman, also of Polygon Labs, Rettig and Mosier argued in a 45-page academic paper that traditional laws and regulations aimed at managing the risk of illicit finance “are not amenable to intermediary-less systems like DeFi.”
Not everything calling itself “DeFi” is truly decentralized. If a human administrator has any level of “independent control” over user funds at any point in time, it makes sense to regulate them just like any other financial institution. But if no one ever has any independent control of the money—as is the case in Tornado Cash—attempting to regulate it like a bank is unworkable, the trio argued.
The major US law designed to combat money laundering is called the Bank Secrecy Act (BSA). As Mosier and Rettig explained on stage at the DC Privacy Summit, the law isn’t a good fit for decentralized financial systems.
The BSA was born in the 1970s out of a simple need for information, Mosier said. Banks frequently didn’t have the records that federal investigators needed to follow a suspected criminal money trail. “So it started with ‘You need to keep records so that when we show up there’s information about which Swiss bank you sent it to, et cetera,’” he said. Eventually, regulators started requiring banks to report to the Treasury’s Financial Crimes Enforcement Network (FinCEN) whenever they observe “suspicious activity.”
The original reason for this is that individual banks are “opaque silos,” Mosier said. Requiring them to share information allows FinCEN to develop system-wide risk assessments, helping the industry stay ahead of the latest forms of identity fraud and other crimes. “We don’t have opaque silos (in the blockchain space). We have an enormous amount of public ledger—very accessible—data,” he said. That information can be used to manage risk.
To open the door for that, policymakers should regulate “genuine DeFi” systems like “critical” communications infrastructure, not banks, Gilman, Mosier, and Rettig proposed. That would place such systems within the remit of the Treasury’s Office of Cybersecurity and Critical Infrastructure Protection (OCCIP). “OCCIP supports the operating integrity of technological and cyber infrastructure that may underpin financial systems as network architecture, without implicating BSA requirements in that security-coordinating function,” they write in the paper.
“There’s all sorts of communications protocols that underlie the multi-trillion dollar financial system,” Mosier said at the DC Privacy Summit. OCCIP collaborates with financial institutions to identify cybersecurity threats and disseminate information so the people maintaining those protocols can respond quickly to new risks. OCCIP plays a crucial role in the nonprofit Financial Services Information Sharing and Analysis Center (FS-ISAC). Some 4,600 banks, credit unions, insurance companies, asset managers, and payment processors participate in the organization, which shares intelligence for system-wide risk management. The crypto industry has already stood up counterparts to FS-ISAC, including the Security Alliance and the Crypto ISAC.
According to Mosier, protecting users from bad actors doesn’t require collecting personal information, the way “know-your-customer,” or KYC, the traditional anti-money-laundering approach, does. “There’s a lot of ways to manage risk that doesn’t just keep the shades off Clearville,” he said, alluding to Van Valkenburgh’s earlier talk.
It’s not just about crypto, Rettig added. “The BSA is not effective anymore,” she said. “It’s not meeting its purpose now and even regulators know that a new digital—forget blockchain on its own—but a new digital system is emerging, and there are better ways.”
In decentralized systems, Rettig and Mosier have proposed that policymakers recognize a new set of entities that, as Rettig explained on stage, “would have additional or enhanced financial integrity requirements.” So-called Critical Communications Transmitters (CCTs) are “people who are necessary to the flow of communications about these transactions,” she said. Their proposal highlights for example “node-as-a-service” businesses that run remote procedure call (RPC) nodes—responsible for receiving communication from individual wallets and then transmitting new transaction information to the network of computers validating those transactions.
Regulators could require these businesses to check “risk scores” for individual wallets, determined via analysis of onchain data, and block transactions from high-risk wallets in real-time. Rettig, Mosier, and Gilman acknowledge in their paper that legislation may be required to create this new regulatory category.
“TLS for finance”
The general idea, though, can be realized without new laws. For example, 0xbow, a startup that grew directly out of the government crackdown on Tornado Cash, is creating cryptographic tools that use onchain information to give users the power to disassociate themselves from known bad actors.
0xbow is aiming to commercialize an idea known as privacy pools. The idea for privacy pools was a direct response to Tornado Cash sanctions, Ameen Soleimani, a longtime Ethereum software developer and founder of an organization called MolochDAO, which provided the Tornado Cash developers with their initial grant funding, explained on the DC Privacy Summit stage. In fact, he said, it was born during a podcast that Ethereum co-creator Vitalik Buterin recorded with Coinbase CEO Brian Armstrong in August 2022, the same month the sanctions came down.
What makes Tornado Cash so novel is the cryptography under the hood, called zero-knowledge cryptography. It allows users to prove that a statement about themselves is true without revealing any other information—hence “zero knowledge.”
The technology was a major topic of discussion throughout the day at the DC Privacy Summit. Zooko Wilcox, who led the team that invented Zcash, the first practical implementation of zero-knowledge proofs, did an on-stage demo of how they work. And I sat down with Justin Thaler, a computer scientist and research partner at a16z Crypto, to talk about where the science of zero-knowledge cryptography stands today.
What Buterin pointed out on that podcast in 2022 was that it is possible to use zero-knowledge cryptography not only for privacy but also to give users of privacy tools like Tornado Cash the power to prove that they aren’t associated with known bad actors. It would be possible, he explained, to create sets of deposits that have been vetted in some way and are known to not be associated with illegal activity. When depositors in those “privacy pools” withdraw, they could use a zero-knowledge proof that verifies their membership in such a set.
“So you would not reveal exactly who you are. You would not even reveal exactly who you are to, like, one specific group,” Buterin said. “What you would be doing is you’d be saying, like, I am some participant in this ecosystem but I am not a hacker.”
Soleimani took that idea and ran with it. MolochDAO funded a prototype and Soleimani convinced Fabian Schär, a professor at the University of Basel, to collaborate with him and Buterin on an academic paper describing the concept. Then he teamed up with Zak Cole to form 0xbow.
0xbow creates what it calls “association sets,” which are simply sets (pools) of deposits that have been vetted in some way, whether by some traditional KYC process or in the sense that the set’s members don’t have reason to distrust each other.
The simplest association set would be like a decentralized, private service to group fund something, like a dinner. It would be like Venmo, only without all the public broadcasting of who’s paying for what, and how much. “I don’t want people to know who’s coming to that dinner necessarily and I don’t want people to know how much they are spending,” Cole said at the DC Privacy Summit. It’s possible to generate an association set and have participants opt into it.
Another example might be a bank consortium or a group of crypto exchanges, Soleimani explained. If you buy from an exchange, for example, it’s clear on the chain which specific exchange it was. That’s not very private. Alternatively, say 20 exchanges created a pool that only allows deposits from individuals who have passed a KYC check at one of the participating exchanges. If someone withdraws from the pool, all anyone would know is that that money came from one of the member exchanges. “It’s a lot harder to invade your privacy that way,” Soleimani said.
Cole said 0xbow is working on some proofs of concept with mainstream financial institutions that are interested in participating in DeFi but want their transactions to be confidential. “This is like TLS for finance,” he said. TLS stands for transport layer security, a cryptographic protocol used in the Hypertext Transfer Protocol Secure (HTTPS), the widely used method for securing internet communications.
“Transaction prerequisites”
Association sets are just one example of how blockchain data can be used to keep bad actors from using decentralized applications or networks for criminal ends. Onchain data can be used in a number of ways to score the risk of individual wallets and deposits, and blacklists of known hackers. A startup called Predicate is creating software tools that use that sort of information to build “transaction prerequisites” into smart contracts, cofounder and CEO Nikhil Raghuveera explained at the DC Privacy Summit.
Raghuveera used a “web2” example to explain the concept of transaction prerequisites: every time you request an Uber, software behind the scenes runs several checks. “It’s saying, okay, where are you located? Who are you? What’s your Uber rating? Where are you trying to get to? What’s traffic? What type of car are you requesting?” he said. “On the back end, it is matching you with a driver that of course meets a number of pre-transaction conditions.”
Predicate aims to help decentralized applications do the same kind of thing, and the first place where this has been useful is in the area of anti-money laundering controls, Raghuveera said. That’s why the company teamed up with Aleo, a recently launched layer one blockchain that uses zero-knowledge proofs to keep transactions private.
The community building applications on Aleo passed a governance proposal designed to ensure that bridges—smart contracts that allowed people to move value from the Ethereum network to the Aleo network—did not let any illicit funds into the Aleo network. It required bridges to perform a number of pre-transaction checks before letting money pass through. Predicate has developed a common mechanism for implementing all of those requirements at once, so that it’s not necessary for every new Aleo bridge to reinvent the wheel, Raghuveera explained.
“Nobody wants to be the bridge that lets in illicit finance,” Aleo’s policy lead Leena Im added on stage at the DC Privacy Summit. “I mean I’m sure some exist, but those aren’t on our network.” The next question is: “How do we then verify that the bridges are continuing to do this? And I think that’s where Predicate and 0xbow and other technical tools come into place.” Aleo is also part of the Crypto ISAC, Im noted.
The challenge is making sure that the process doesn’t introduce any single points of failure “We’re only a month old,” Im said. “So until a very robust community of independent, compliance or risk-mitigation focused devs pops up, like, how much do we weigh in as a foundation? How much might Predicate or other service providers weigh in once they are part of the ecosystem?”
The inevitability of the technology
Questions like these reflect how the nexus between human governance and decentralized, automated rulemaking is still evolving. On the technology side, at least, that evolution is inevitable. That was the point of Van Valkenburgh’s parable.
Clearville is Bitcoin—and Ethereum, which inherited Bitcoin’s transparency. Shade is the Tornado Cash developers. He didn’t invent curtains. In fact, he told the Clearvillians of other visibility-killing materials that existed in his homeland, Unclear. “There was paint, wood, and stone—even vinyl veneer.”
Putting Shade in jail clearly won’t stop the covering of glass walls. Shade technology feels inevitable. Besides, Clearville’s houses were made out of glass because, before Shade showed up, it was the only building material available.
“We should not assume that the inventors of incremental improvements are inherently doing something wrong—especially in the absence of their clear intent to assist known criminals,” Van Valkenburgh added. Private decentralized systems feel just as inevitable as shades. “If not Tornado Cash, then there are countless other projects and potential protocols which will continually improve, bringing privacy not just to Ethereum but to Bitcoin as well,” Van Valkenburgh said.
“At what point is someone doing more than sharing tools for general usage? …At what point are they conspiring to further some crime by providing those tools?” Van Valkenburgh asked the audience. “If they have no personal interaction with those potential criminal users beyond publishing a smart contract and a website that criminals, along with others, use, are they culpable?”
These questions are not rhetorical. It might be possible to remove all the shades from Clearville. But blockchains make tools like Tornado Cash virtually unstoppable. “It is likely a fool's errand to forbid anyone from using privacy tools online.”
What are the alternatives? The DC Privacy Summit showcased a new species of tools designed to filter out the bad actors from the good ones. Is that approach fair? Is it Constitutional? “Who decides which uses of the privacy tools are good and which are forbidden?” Van Valkenburgh wondered.
Finally, in the context of US law, who would craft all these overarching policies? Existing laws haven’t been clear enough, and that’s led to fighting in the courts. The most relevant example here is the prosecution of Roman Storm. Regulators at FinCEN and the Department of Justice don’t seem to agree on whether Tornado Cash is a money transmitter—and thus whether it should have been required to register with FinCEN. And the letter of US sanctions law suggests the Treasury has overreached in its sanctions of the protocol, Van Valkenburgh argued. “Can we criminalize behaviors that are essentially new under existing laws or do we need to draft new laws and ensure that people know and understand those new laws before expecting them to comply with them?”
“I don’t want to live in Clearville. I don’t think you can force me to live in Clearville if I’ve not committed other crimes,” concluded Van Valkenburgh. “And I don’t think you can arrest people who give me tools for privacy in my entirely legal day-to-day affairs, merely because someone else who got the benefit of that privacy wanted to break the law.”
“And, even if I have to live in Clearville… I would at least expect the dignity of there being a public debate amongst my elected officials, a vote in Congress, and a signature from the President before they pull down my curtains and stare through my walls.” —Mike Orcutt
Follow us on Twitter or get corporate with us on LinkedIn—if you want.