A deep dive into decentralized autonomous organizations and their promise to dismantle traditional governance structures
Decentralized autonomous organizations are fast becoming the favored governance model of the crypto-native. From decentralized-finance protocols to collectors of digital artwork and legal experts specializing in crypto-related matters, distributed decision-making is evolving and threatening to make traditional, top-down organizational models obsolete.
The latest protocol to pass control to its users is the Ethereum Name Service — a project that links human-readable names to the more accident-prone and less visually appealing wallet addresses. To facilitate community governance, Ethereum Name Service, like many other protocols, uses token ownership to grant holders voting rights on the protocol’s future direction. Platform users were invited to claim ENS tokens in early November and use it to vote on the new DAOs constitution, treasury control and all matters going forward.
This OKX Insights in-depth article looks at the evolution of distributed governance mechanisms, as well as their strengths and weaknesses. Ultimately, we ask if the decentralized decision-making processes of the many DAOs emerging today look set to challenge those traditional, hierarchical corporate structures found outside the cryptocurrency industry.
What is a DAO?
At its simplest, a DAO is a distributed group of individuals who unite to forward a common interest or goal. That might be investing in digital artwork, managing the rewards paid out by a decentralized-finance protocol or providing a service to exterior parties.
“Every DAO will have a different core mission, in the same way that every company has a different core mission. DAOs are basically a way to coordinate like-minded individuals to solve some common goal.”
Outside of the blockchain industry, typical organizations favor a rigid, top-down decision-making approach. Startups, for example, often have a fixed hierarchical structure with a chief executive officer at its head. The CEO’s word is usually final, even if their approach is not best for the wider organization. In such a structure, legal contracts enforce the relationship between members.
In a DAO, smart contracts take the place of legal contracts. Deployed on blockchains like Ethereum, these self-executing digital agreements inform the scope of community decision-making. For example, in a protocol like MakerDAO, MKR holders can vote on variable protocol parameters such as the Stability Fees required to stabilize the DAI stablecoin’s USD peg. Increasing the fee and, therefore, the cost of borrowing DAI should reduce new demand and raise the token’s market price.
DAO members can submit proposals to adjust governed variables using votes cast with the project’s governance tokens. Should a proposal receive enough votes to pass a previously agreed-upon quorum, the change is implemented automatically via the smart contract.
Evolution of DAOs
Although the teams behind many blockchain-based protocols have only just started handing control to their communities through DAO governance, the concept of decentralized decision-making is not a new one. Indeed, Ethereum cofounder Vitalik Buterin wrote extensively about distributed governance systems as early as 2014 — the same year that Rune Christensen first outlined the Maker protocol, its decentralized stablecoin and community leadership. In explaining the concept, Buterin stated:
“The ideal of a decentralized autonomous organization is easy to describe: it is an entity that lives on the internet and exists autonomously, but also heavily relies on hiring individuals to perform certain tasks that the automaton itself cannot do.”
Unfortunately, Ethereum’s first major DAO was anything but a success. Confusingly titled “The DAO,” the experiment in collective investing took funds from contributors into its treasury in exchange for DAO tokens. These tokens gave stakeholders votes in the distribution of funds and a claim on any profits generated.
During its 2016 launch, a critical smart contract bug was exploited in The DAO, enabling those responsible to drain around 3.6 million ETH from the collective fund. The controversial decision was made to roll back the Ethereum blockchain to a state from before the incident, returning all compromised funds to their previous owners. The fledgling Ethereum community was divided on such a response, with an opposition minority choosing to continue without the rollback. The resulting hard fork created the Ethereum and Ethereum Classic networks.
In the wake of The DAO exploit, public enthusiasm was somewhat dampened for decentralized governance. However, not everyone dismissed the concept. Notably, developers at the Aragon Network took on the task of building DAO tools that would not expose its users to the same protocol risks that mistakes in The DAO code invited. Having concluded an initial coin offering in 2017, Aragon launched on the Ethereum mainnet in 2018 and began to introduce various products to simplify establishing a DAO. In 2020, Aragon developers handed control of the platform to its community as it became a DAO itself.
Aragon’s product suite provides many of the features a successful DAO needs to organize itself and has become one of the most popular ways to establish a DAO today. Although far from a complete catalog of every organization describing itself as a DAO today, many of the projects listed on DeepDAO.io use some of Aragon’s tools to achieve decentralized governance.
With most blockchain development being open-source, recent years have seen other DAOs emerge, and protocol teams borrow their structures for their own governance. An early example was the nonprofit Ethereum development fundraising group Moloch DAO, which was forked by the for-profit investment DAO MetaCartel. Similarly, another massive name in DeFi, Compound, developed its own decentralized governance infrastructure — and in 2020, its CEO, Robert Leshner, encouraged other teams to borrow from it:
“Compound created a basic governance token and voting framework which we hope other teams adopt to accelerate their own development.”
Indeed, speaking on the Bankless podcast, Nick Johnson, the founder of Ethereum Name Service, explained the rationale he and his team took for establishing a DAO using existing models:
“I really like the Compound-style contracts for on-chain governance, with the delegation setup and so on, because a lot of it requires on-chain voting and the delegate system reduces the number of people that have to vote and it gives you the highest level of technical security against manipulation that’s available.”
Johnson went on to comment that the ENS DAO uses Open Zeppelin’s tooling, which itself is heavily inspired by Compound’s governance model. He added that Gitcoin’s structure was also an influence when designing the framework.
These pre-built DAO structures certainly reduce the technical understanding required to implement decentralized governance systems. However, as Johnson alluded to, outside of minimizing the number of on-chain votes, they do not address one of the major obstacles encountered by smaller groups establishing DAOs — i.e., Ethereum gas fees. Occasional on-chain governance votes relating to massive protocols like Maker and Compound might be worth the transaction fees required to cast a vote with MKR or COMP; however, requiring users to vote by transferring governance tokens is an ill fit for upstart or less financially motivated DAOs.
In 2020, developers from the DeFi protocol Balancer launched Snapshot. The off-chain, gasless voting client enabled members to submit proposals and have DAO members vote off-chain using a project’s governance tokens. Initially, Snapshot was entirely off-chain and required trust in a DAO’s core development team to enact the results of a vote. However, a recent integration with decentralized multi-signature solution Gnosis Safe linked Snapshot votes with an organization’s treasury — enabling the direct spending of on-chain treasury funds via voting.
Advantages of DAO governance
With projects like Aragon, Moloch and Compound leading the way, and tools like Snapshot reducing overall costs, community governance as a concept has taken off in recent years. In fact, it is becoming more and more common to see newly launched tokens include governance utility as one of their intended applications. Given several key advantages over traditional organizational structures, such a trend is likely to continue — at least for the crypto-native.
Nick Almond, the protocol leader at DAO application suite finance.vote, touched upon some of these benefits in correspondence with OKX Insights:
“I think there is the potential for DAOs to be radically more open and digitally native than conventional organizations. They are closer to digital democracies than companies. They are truly global, can tap into massively international markets, are talent-based and are implicitly crypto-centric (which conventional organizations most certainly aren’t).”
Mentioning to OKX Insights that distributed ownership and power was key among DAOs’ advantages, Spencer Graham, a core contributor at infrastructure provider DAOhaus, elaborated:
“A few high-level benefits of distributed ownership and power are: more equitable distribution of power, better information gathering and processing capability, decisions reflect the preferences of the entire community, and the right leadership can emerge from anywhere. But it also means that there’s less trust required among members, so DAOs can attract top talent because they offer much greater autonomy and flexibility, and it is also much easier for new contributors to start adding value quickly.”
Another major advantage of DAOs when compared to traditional company structures is their transparency. The code upon which the DAO operates is open-source, meaning that anyone is free to check how a group functions before participating.
In addition to the structure itself being transparent, members are identifiable by the tokens they hold on-chain and votes take place publicly. Combined with multi-sig treasury wallet implementations like those provided by Gnosis, these factors reduce the likelihood of managerial corruption impacting a DAO’s performance.
Finally, for some of the protocols adopting it, DAO governance is just a more natural fit. Ethereum Name Service, for example, is intended to be a public good for all internet users rather than a for-profit enterprise. Given its distributed appeal, ceding control to the platform’s users seems appropriate. When asked why the decision was made to pursue DAO governance, Johnson told the Bankless audience:
“If we’re building a system that should be useful to a wide variety of people, we shouldn’t be relying on one single person’s judgment on exactly how that should operate and what changes they should make. […] We need a reasonably decentralized system with a variety of viewpoints to decide what’s best for the users. And I think the best way to do that is to actually have the users involved.”
Obstacles facing DAOs
DAOs and the law
As international, decentralized organizations, DAOs do not fit neatly into existing legal frameworks. A typical company will be registered in a particular jurisdiction and take legal responsibility for any liabilities and obligations it might incur. It is much less clear who should be held responsible in the case of a DAO — which is typically not registered as a legal entity anywhere.
According to a report by legal firm Withers Worldwide, without prior registration to the contrary, United States law classifies DAO participants as general partners. Should the DAO incur a financial liability through its actions, this would require every token holder to cover the expense. The same sentiment is expressed in an anonymous post detailing the options for legally structuring a DAO. Having apparently spoken to several legal experts, the author presents five options, each of which has distinct disadvantages.
The first option is not to establish a legal entity for the DAO at all. Calling this “a bad idea,” the author states that all DAO members can be held fully liable should a case against the organization be brought to court. The author then mentions the tax treatment of unregistered DAOs, arguing that individual members could owe taxes on DAO profits, even if they themselves had not personally made any money from their interactions with the organization.
Their final point addresses real-world interactions. While not strictly relevant for all DAOs — as some deal exclusively in the digital realm — those that want to sign real-world contracts, hold non-crypto assets or own intellectual property cannot legally do so without first registering as a recognized entity.
Some options are emerging for DAOs seeking to register as legal entities. In various jurisdictions — for example, Wyoming and Delaware — DAOs are permitted to register as a limited liability company. This extends some legal protection to members in that any liabilities or obligations incurred are the responsibility of the DAO as a whole, not its individual members. Other options include registering as an onshore or offshore foundation — or some combination of the above. However, the taxation issues surrounding tokens issued later being used to provide liquidity in a protocol could make registering as a foundation somewhat unattractive.
Speaking to OKX Insights, Almond added:
“The definition of what they [DAOs] are is so fluid right now (and might always be) that I’m skeptical it’s possible to provide an expressive enough legal wrapper to facilitate the kind of innovation that’s happening within them at the moment. I think DAOs that want to do things like hold real-world assets should be exploring this, but truly digital asset DAOs are probably better staying that way for the time being. Also, is a DAO autonomous if it’s picking a jurisdiction and complying to its laws and can be stopped by them?”
Graham expressed a similar sentiment to us, reasoning that the differences between today’s DAOs make their legal status tough to define:
“There are so many types of DAOs, many of which use different structures. Some DAOs (such as those on DAOhaus, which use the Moloch DAO framework) map pretty well to a traditional LLC structure. Some DAOs fit better with a cooperative model like the Colorado LCA structure. And other DAOs don’t fit well with current legal frameworks.”
DAOs and the regulators
Another issue that may impact certain DAOs is their token’s treatment by global regulators. The U.S. Securities and Exchange Commission has moved somewhat tentatively in attempting to police the expanding cryptocurrency ecosystem. However, it did posthumously declare that The DAO’s tokens should have been classified as a security. As such, their issuer — The DAO itself — should have been registered with the regulator.
The issue of whether other DAO governance tokens should be classified as securities appears to come down to whether or not its founders are essential to the enterprise and, thus, its profit potential. In the case of The DAO, a central party, slock.it, was integral to the drafting of documentation relating to the organization, maintaining online forums used to communicate and answering questions from prospective investors.
When addressing The DAO members’ involvement relative to slock.it, the SEC writes:
“Although DAO Token holders were afforded voting rights, these voting rights were limited. DAO Token holders were substantially reliant on the managerial efforts of Slock.it, its co-founders, and the Curators. Even if an investor’s efforts help to make an enterprise profitable, those efforts do not necessarily equate with a promoter’s significant managerial efforts or control over the enterprise.”
The precise makeup of a DAO, therefore, influences its governance tokens’ classification with the SEC. While those projects like Ethereum Name Service may well have passed sufficient control back to users, this is surely not the case with every organization using the term DAO to describe itself today.
Should the SEC bring an action, such a group’s legal classification would become even more of a pressing issue. Tokens previously classified as securities by the regulator have seen their issuers receive significant fines. As mentioned previously, without existing as a legal entity, DAO members themselves could be held personally liable for similar penalties.
DAOs and centralization
Although their naming suggests a lack of a central point of failure, in reality, some groups using the term DAO today show little meaningful decentralization. A recent example was the AnubisDAO, in which a single signature account was used to deploy its initial token sale on the Cooper liquidity bootstrapping platform. This enabled a lone user — whether that be founding member “Beerus” or, as they claim, a third party compromising their system — to drain around $57 million from the project before it had even launched.
Even in more established projects, centralization risk is prevalent. In a recent OKX Insights in-depth, we looked at Olympus DAO. While its sizable treasury resulting from its innovative liquidity purchasing mechanism is impressive, it remains under the control of just four of a possible seven Gnosis Safe key holders. Even in the event that all seven individuals are forever loyal to the project, it is not beyond the realms of possibility that an attacker could discover the identities of the key holders and attempt to extort them. After all, the treasury currently holds more than $500 million — a sizable payday.
DAOs and token voting
While many projects have been quick to embrace governance token votes — particularly as tooling via Snapshot, Aragon and other platforms make adopting decentralized governance structures more accessible — several influential observers remain critical. Among them is Vitalik Buterin.
In an August 2021 blog post, Buterin outlined several issues with current tokenized governance structures. Among his issues with the model is that coin holders themselves are central to the governance process. This creates a potential conflict of interest, an example being when a particular decision may hurt the protocol long-term but could result in a higher price in the short term.
However, Buterin’s main contention with current token-governance models is that votes can be bought. He uses the example of a simple token-wrapping contract that rewards participation and that auctions voting rights to the highest bidder. Reasoning that a small-holder is likely to accept such a proposition, regardless of consequences to the protocol itself, Buterin writes:
“If the governance power auctioned by the wrapper contract gets bought up by an attacker, you personally only suffer a small fraction of the cost of the bad governance decisions that your token is contributing to, but you personally gain the full benefit of the dividend from the governance rights auction. This situation is a classic tragedy of the commons.”
Although such attacks are rare today, the Ethereum cofounder believes they are coming. According to him, community spirit and a lack of tools to orchestrate vote-buying may be keeping threats at bay for now. He concludes by stating that the Ethereum community should not consider coin-voting mechanisms in action today as “safe defaults” because few have been tested under great economic stress. Instead, developers should be actively exploring alternatives.
The organizations of tomorrow?
The availability of improved tooling from the likes of Gnosis, Aragon and others has made DAO structures more popular than ever. Always a neat fit with the censorship-resistant, trust-minimized and self-sovereign ambitions of the crypto industry, the technology required for community governance has improved to the point where incidents like The DAO should be rare.
Contrary to the corporate world’s traditional top-down governance structures, DAOs look to grant control to the people with the most interest vested in a blockchain-based project. Such democratization of leadership is certainly a commendable goal and provides a platform for many voices to put forward ideas about future development.
Owing to their advantages, Graham believes that most organizations will eventually realize the benefit of decentralized governance. He told us:
“I do expect that at some point in the not-too-distant future, nearly all organizations will be structured as DAOs. Some of that will be existing companies that have transformed into DAOs. However, I suspect that most of the transformation will be organizations that start as DAOs outperforming and eventually obsoleting existing companies that resist transformation.”
However, a more cynical observer may accuse those forming DAOs to govern protocols that may threaten the global financial status quo of washing their hands of responsibility. Traditionally, those closest to an organization are legally liable for it. However, in a DAO, that control is passed wholesale onto a new, distributed organization when registered as a legal entity or those participating members when no such entity exists.
In addition to their legal and regulatory treatment, other issues make DAOs in their current form less attractive. Buterin’s points about the inefficiencies of token-voting models may eventually threaten the ever-increasing number of DAOs using this structure currently, and centralization risks threaten naive users rushing to participate in such organizations.
Adding to the factors mentioned above, the protocols transitioning to decentralized governance remain relatively immature. Illustrating this concept neatly is the recent example of the Compound protocol.
As highlighted by Johnson and those other protocols adopting it, the Compound governance structure is one of the industry’s most established and admired. However, in the event of a catastrophe, even such a well-designed governance structure has issues. When Compound recently shipped an update containing exploitable code, no entity was equipped with sufficient control to enact any change without going through a seven-day governance vote. Meanwhile, the exploit caused around $162 million in COMP tokens to be put at risk.
Indeed, owing to the various issues faced today, Almond believes that DAOs will remain the domain of “digital mavericks” for now. However, he does foresee a not-too-distant future in which the line between traditional and decentralized organizations blurs:
“I think we’ll see the emergence of hybrid decentralized organizations, where fairly conventional organizations will inherit practices from the DAO space that have been proven to be useful. They’ll be great for international decentralized workers and when traditional organizations start playing more deeply with crypto assets and want to loosen their organizational boundaries and let outside expertise in. I think we’ll start seeing some progressive organizations do this fairly soon.”