[Signal Request] Should MakerDAO Assist in recovering DAI locked on Optimism Escrow?


Pull lost funds manually from the optimism escrow via an executive vote.


A user accidentally called the burn function on the DAI.sol on optimism, accidentally burning 10,000,000 DAI.

Here is the transaction hash showing the event: https://optimistic.etherscan.io/tx/0xd2a83ab5e9c412a5ce9e738e967051ebef9b92459a5732858a06cdf82ba469c8

The error occurred given that the burn function was confused with that of this contract: https://optimistic.etherscan.io/address/0xad32aa4bff8b61b4ae07e3ba437cf81100af0cd7

Given that it is easy to argue this was not intentional (nobody would send $10M USD to a 0 address), this post kindly requests to recover this funds via executive vote.

Protocol Engineering has confirmed that it is possible to do this on-chain by pulling the funds manually from the Optimism escrow via an executive vote.

The user can include a signed message with the wallet “I called the burn function in the DAI contract out of confusion with a different contract, please help me recover the funds”, if such proof is required.

Hopefully the community will understand that this was a human error and help recover the funds.


Should there be an executive to manually pull the burnt 10 Million DAI from the Optimism escrow via an execution vote?
  • Yes
  • No
  • Abstain

0 voters

Next steps
This poll will close on January 7, 2022


Just to cover all the Signal Request requirements @lost_some_dai , the end of this poll should include how long voting is taking place. The standard there is 2 weeks, which I might suggest extending slightly due to our holiday break but that decision is yours.

GovAlpha can add the next steps to your post once you pick a conclusion date for the Signal.

1 Like

Thanks for your reply @prose11 I thought I had followed the steps correctly from this Guide to Signalling.

I opted for what was indicated there, and crossed the x in “automatically close poll”, as pictured below:
Screenshot 2021-12-20 at 16.26.13

Is that the parameter I should edit? Thanks!

Hey friends, it’s been a while :slight_smile:

For some reason I cannot vote on the poll, but I would vote yes. It’s an incredibly large sum of Dai, on an L2 we’re all still adapting to use, and it’s the right thing to do.

Thank you!

(picture of message saying I need to be level 2)


The poll parameter looks to be correct, but your post should indicate at the bottom when the vote is closing. That way people know how long they have to discuss and make up their minds on this issue.

So two weeks has this poll going until Jan. 3rd, unless you’d like to extend it further.

Thank you. Made the edit and added a few days as you suggested.

Please let me know if there are any changes you would recommend.

1 Like

Hi @lost_some_dai, at some point we will ask you to publicly sign a message to confirm ownership of the affected wallet. Preferably this would include the forum username you used for this post.

You could do this now if you like, or wait to see if the Signal Request is successful first, but we will request that this is done before we put a vote on-chain.

We will ask @Protocol-Engineering to confirm the contents of the message but the community should also be able to verify for themselves, since we will be asking them to approve a transfer of 10 million DAI.


I voted no, since it looks quite arbitrary without more information . While this is large amount, can you provide more details for reasons to act very fast? I assume answer would be Maker/Dai interests on L2. But, still. Things should be explained, not just: “we are buddies with common interest, let’s do a quick executive vote”.

Would we do such thing for 1m? for 300k etc.?

Just to be clear, I have no problem with recovering this amount - it makes sense.
I would also expect new user ‘lost_some_dai’ would introduce himself.


I feel like this pool should have more options like: yes but with a [FEE]%.

It’s still a strong YES from me. We can do this, and we should do this but DAO should think about this problem more holistically and maybe create a “Customer service” CU.


I definitely agree with the fee part. It could be set as a percentage of the amount being recovered with a floor value or simply a fixed amount.

At the bare minimum, it should cover the PE time spent on this but in general, should be more to avoid the protocol having to spend engineering time to bear the cost of user error.


Hi @Patrick_J, thanks for the information. Here is the signed message:

“address”: “0xc9b48B787141595156d9a7aca4BC7De1Ca7b5eF6”,
“msg”: “I am lost_some_dai and I am signing this message to verify that I am the owner of the address that accidentally burnt 10M DAI”,
“version”: “2”

Please let me know if this suffices or if you would need something else. Thanks again!


Message signature verified per Etherscan.



Thanks Brian.

@lost_some_dai that should be sufficient. We’ll let you know if we need any further information.

1 Like

@brianmcmichael @Patrick_J Thanks a lot!

1 Like

So I’m going to be a wet towel, but defer to Engineering and others and hope they answer - and why I abstained rather than voted no initially because I acknowledge others know more than me. I sleep well not knowing who or what group is using the protocol to borrow DAI against collateral, because we aren’t actively assisting anyone and it is part of the fact the protocol does not discriminate based upon race, ethnicity, sex, etc. But here when we are asked to do an affirmative act for someone to retrieve an asset the value of which is $10 million, don’t we have an obligation to know who is doing the asking and where they are from? If someone on Interpol’s red notice list buys a soda from a vending machine I own that is one thing, but if that same person knocks on my door and says help me retrieve $10 million I accidentally dropped in the storm sewer drain next to my house isn’t that quite another and can’t I get in trouble if I do so help that person?


While the analogy is interesting it ignores a little too much of the specifics of cryptographic proof and the fact that it is possible to act neutrally here. That being said the parallel discussion should yield a long-term solution (say “Every 6 months all Dai sent to the Dai contract and the 0 address are made available to their owners at once”) that avoids these issues and just adds a kind of self-repair mechanism to the Dai machine.


It is with a heavy heart I voted “No” in this poll. Let me immediately qualify that by saying I do believe we should help recover the funds, but the way this poll is structured it seems to be a straight up recovery with no strings attached.

As most are probably aware there is a proposal about instating a policy regarding recovering lost DAI. I believe we need to treat all users equally and fairly and thus I believe we need to instate such a policy first before we start recovering any funds. I am aware that forming such a policy is going to take some time but it seems to me to be a necessary first step before we commence a recovery operation.

There are many who have lost DAI through user error in one form or another and others would, in my opinion, rightfully be angered if we select some to be recovered and not others.

I ask for patience while we get a policy in place and promise I will do my part to help in the recovery once we have done so.


@lost_some_dai In planning this for spell execution, we need to know if you prefer to have the Dai returned on L1/mainnet or if you are looking for it to be sent to the address on L2. It would be less complicated both from a technical perspective and an accounting one to handle the transfer on by pulling from the optimism escrow on mainnet.


1 Like

It is fine to have it on L1. Thanks!

1 Like

Hi Justin. Thanks for commenting and voting.

Though I understand the nature of your arguments, I think there are cases where the magnitude of a situation matters.

Happy to follow through on the discussion for a formalised procedure in the future to help additional users that had any issues, but hope that the magnitude of this specific incident is enough to persuade you to vote yes on the poll and that this serves as a small precedent to showcase the community that it is possible to assist users in recovery.