Governance Security Beyond Code Navigating Political, Social, and Economic Attack Vectors

[Music] all right we are back ladies and gentlemen we are and now we have Alex and then Daniela with block for please give a round of applause to these guys here guys hello how are you all I hope you guys are enjoying it Denver um well let's talk about governance security Beyond code like we have seen like contract out dates uh for a long time but we don't talk much about governance security like delegation rates slow rugs collusions bribes so we are going to navigate political social and economic attack vectors in Dows how can we protect the Dow ecosystem how can we prevent the Dows to be drained how can we make Dows um create longevity for Dows um well I'm Danny I'm from Brazil I'm product manager AB Blu uh also delegate representative in arbitr scroll and part of balancer up Co I'm going to pass the mic hello everyone I'm Alex uh Co of block also ens met governance uh Steward and yeah let's get into it um well about blogfull we are our goal is to prevent governance attacks uh creating Cutting eding Edge research and Technology to prevent this well where where it basically started um so in February actually 2024 uh we are we were already active in ens governance and and so by being active we we saw the numbers of uh the amount of tokens that were participating on governance which were basically 4 million uh out of like 100 million total Supply 33 million circulating Supply only 4% was delegated and inside uh for these four perent these four uh million only 1.5 1.5 million of these tokens were actually participating in a governance so imagine that like a a company like with shareholders right and only 1.5% of the shareholders are actually taking decisions and and so with that we started questioning ourselves what what are the consequences of that like what does these numbers um means what are the risks and and so this started uh quite interesting research um so why why governance security um as you can see in this chart um the blue line is basically the assets that the Dow has in eth and USD so these are liquid assets you can like it's it's different from the governance token right the governance token has an amount of liquidity you cannot sell everything so this is this these are liquid assets uh the blue line and then the orange orange line is the value of all delegated tokens so basically all the tokens that participate on governance and and so with the research we we have done we uncovered this uh critical scenario of vulnerability inside the ens uh Dow which could have led to 150 million uh Taft at one point and this sounds like a theoretical attack but it's actually something that is quite common and in in D and attackers in the past have done much much more complex attacks for like lower values so it's actually like a pretty real risk and not just that like the economical um vulnerability is actually just one part of the attack the other one is uh the governor implementation and so this was quite critical the one block voting delay what this means is basically the attacker with enough uh voting power he could basically propose the vote and the vote that could move the whole treasury to himself and one block later he called vote in this vote right and dump the tokens he don't need the tokens anymore because his vote was already casted and so it could be be really like an an instant attack or quite quick attack with just one block uh of delay um so it's really really hard to defend against that because token holders they don't have much time to to react to that they don't have much time to delegate tokens to another delegate or to themselves so they can vote against this malicious vote so one one block voting delay it's something quite critical on on attack and and so you see it's it's not only the economic vulnerability but it's also the the governance implementation itself one one other point is the cancel function um so in the uness governor there there there was no cancel function uh which is also not helpful uh and so the ideal here is that after a malicious proposals come in you could somehow make another voting for can uh the malicious proposal right like a normal voting but um that's also uh not not not enabled um and so yeah once the The Proposal was live once the proposal um yeah once the proposal is live there is basically no Counterattack uh you could do to defend and so economical vulnerability plus weak uh Governor implementation or or at least Governor implementations that have these these type of of characteristics it's like a really easy attack it's like a a vulnerable situation to be in and and so the midterm solution that we proposed was the security Council uh so after identifying this this the scenario um we proposed the security Council which is basically uh four out of eight uh multisig that can uh veto uh any type of of uh proposal but obviously only malicious proposals so that's that's subjective but in the code uh this this the security Council can V uh any any proposal and then the understanding of what's malicious it's of course subjective because you cannot enforce on the code uh also because if you see um some other attacks like the compound attack the last year uh it's quite subjective right like people some people say it wasn't an attack other people say it was um so yeah that that like governance attacks they can be uh quite subjective uh this is a bit of more detail of how uh the security Council looks so we don't want the security Council to have uh a power for a long time or like make them um let's let's for example do a proposal for so now we don't need the security Council anymore right if they can veto this proposal they stay in power forever so one quite quite important aspect is that anyone can revoke the security uh the security Council power after two years and and with that mechanism we ensure that we don't concentrate the power too long on the security Council so we need to come up with a solution uh for one one one and a half year more because this were like this was six months ago and and yeah this is the current state so last year when this proposal was live this was the situation of uh delegation on the Dow so liquid treasury again the treasury in eth and USD was 11020 Million the value of all the delegated tokens which is delegated cap was 65 million and the value of the average turnout uh the the the tokens that usually participate on on voting they are they were around 24 million meaning that if you use the average turnout as a metric you could basically do a a 5x profit right it's it's like crazy but then other mechanisms can also be used to lower the cost of this attack um so some examples of that is if you know you are attacking the Dow and it will be a successful proposal you can also short the token in exchanges you can short the tokens on Futures you can also uh just do some incentives for getting delegations because all you need is delegation you you you need the power you just you don't need a token you need delegation um and other aspect is maybe you could go to Avi and borrow some tokens as well so you don't have economical skin on the game right you you you just borrow some tokens and and use their power to submit this proposal um so for sure you you you can make the attack cost uh lower and lower uh the more different mechanisms you you use uh and so this this led to uh a longer research um what more than 30 Dows and and this keeps growing because I mean smaller Dows they are easier to attack and we cannot track all the Dows so uh we met we M at more than 30 attacks in Dallas and what we learned uh these are like the main learnings are that attackers they are really really creative like it's so crazy um an example I can give is um some attackers bought tokens from uh FTX after the crash um through uh an auction and they use it the tax from like it was like a mango market like solenad to approve an a proposal to Reby the tokens for a higher price from them so they basically profited on top of the Dow making participating on an auction from a bankrupted sex like look how crazy it is like so attackers they can be really creative we cannot like uh sub judge their their creativity and intelligence uh normality bias is more present than in smart contract security so we are more used to invest on Smart contract security risks and and and audits for feeling safe on on on Smart contracts but on governance we have a lot of normality bias which means basically um it didn't happen to me yet so it will not happen right that that's a cognitive bias is really common uh and that's really present in governance attacks and people people think it that governance attacks are only about a 51% attack I mean it is but it's just like a small partion you need to analyze much more than 51% uh tokens and attack to to have an an effective risk assessment of different scenarios so that's that's just the surface thank you uh so what are we doing uh with all those problems we have seen in governance basically we are acting D doing this governance uh security oriented approach we act uh nowadays in ANS Arbitron we are acting as well in Unis Foundation shutter Network optimism scroll as delegate Services providers but also as a guarantee as you can see UNP Foundation announced like a nice grants for governance security um yeah go ahead thank you um why is this so important is that since Dows exist um already 400 million have already been taken from the market just through governance attacks and I'm not including the Dow here the Dow was a code attack we are including governance attacks through politicians politics meanings governance collusions bribes and so on and our purpose here is to defend the ecosystem uh if it had exist a few years ago this four 100 Millions will be here in the market right now like boosting empowering Builders uh within the ecosystem uh and I'm not talking about like a high value here for the market like in comparison with we all have in the market for sure 400 million is not that much but we would be have we would have fewer people questioning the legitimacy of the D ecosystem right now and as you can see people are bearish right now unfortunately and what are we doing so how does anti capture Works uh block food developed anti- capture which is a service and also a product uh we do research and also we are building a dashboard is not live yet but soon we will be live so you can see all the data from the and this research basically analyzes all the aspects and particular aspects of each Dow we know that we have different governance different governance models we have different way of treasures different way of delegations and delegates as well so we analyze even how delegates uh act in in each proposals here is a bit of uh what we analyze it's a it's a bit small I know but we analyze DNS collusions bribes if the the the the Dow has like a meta protocol as well um which kind of front end they use to vote um liid delegation delegation systems and there's a bit of our uh dashboard as well we we have governance activity token uh sex Supply deck Supply so everyone can see how the Dow is going and very soon you're going to have de extractable value so the Dow the the head of governance can see uh how much it costs to attack the da which is very interesting because they can see before the da being attacked and what they can do to prevent that here's a bit of our principles so how we can make profit by attacking this Dow is our first thought but we are red team hacker principles we have that so we are going to be offensive try to attack the da but as we are white hat hackers we are not going to do that we're going to know how to attack the da we're not we are not going to attack meaning that if the da offers us $6 and we know that if we attack the da we can have a 100 million we are going to accept $6 in order to not attack the da and save the da this is uh one of our biggest uh principle here white hat hackers uh meaning that our our principal relies on Save the Dow we realize on we realize towards the longevity of that Dow thank you yeah here's a meme love of memes and um here how we helping the ecosystem uh mainly The Ether ecosystem for sure um providing Dows with actionable steps so not we are not only going to give the Dows the data here's the data we are going to be attacked soon take care no we are going to give them actionable steps hello you need to increase your delegation rate you need to find meanings to do I don't know delegate incentives in order to have more delegates active in your ecosystem govern Governor this second is very very interesting is that we are creating stages like L2 beat we are going to be an L to beat but for governance security Define in stage zero stage one and stage two um so we can measure governance security and create a a standard and I structure safe path for DS towards security prevent marketing losses I already mentioned before about the 400 million so we going to prevent that and the Dow will be safe to invest on Builders invest on their own ecosystems and working towards D longevity by preserving the treasures and their ecosystem here's a nice meme I did I don't know if I if I'm able to have Lazarus hackers but this is our goal and here's our Twitter if you want to get involved it you can be our early follower uh anti capture Twitter is already live uh by following us we can receive like links you can get in touch with me as well uh text me on telegram if you want to be a tester uh very soon you're going to open the an capture uh dashboard so you can see all the data give me feedbacks we are in that phase that we need feedback to improve the dashboard and we are going to have something very very nice that you can be able you'll be able to support So if you are part of the Arbitron Dow or I don't know scroll down you'll be able to vote to sign a petition uh to show that you want that down there on the dashboard included in order to see those datas and stay calm and sleep very well knowing that the treasury not be stolen so that's it that's a this thank you so much [Applause]