Hi everyone, thanks for joining the call yesterday! For those who couldn’t make it or would like to revisit the discussion, here’s the recording: Meeting Recording
As discussed, we’re moving forward with an optimistic governance approach regarding when proposals can be canceled. Our recommendation is that proposals should only be cancelable by the proposer during the pending period before voting begins, unless the proposal is clearly malicious. This is to ensure the integrity of the voting process and respect the legitimacy of votes already cast.
We’ve suggested a two-week contest period, ending on August 4, 2025. If no strong objections are raised by then, we’ll consider this approach approved by the DAO. After August 4, we’ll evaluate with other key stakeholders like the Foundation on whether this update can be bundled with other constitutional AIPs for an onchain vote. If so, we’ll move forward with launching the proposal on Tally.
Let me know if you have any questions. Thanks again!
Hi everyone! Since we have not received any feedback or strong objections on when the proposal cancellation function can be called in the 2 week period, we are proceeding with the DAO’s optimistic approval that proposals should only be cancelable by the proposer during the pending period before voting begins, unless the proposal is clearly malicious.
We will work together with the @Arbitrum Foundation moving forward on next steps, and update the DAO when this is ready to go onchain via a constitutional AIP on Tally. Thank you!
Paulo raised a good question, and I believe we don’t even need to answer it directly — we can simply remove this clause.
After all, we already have another tool in place: the Security Council.
That’s exactly what this committee is for — to halt malicious transactions (as defined by the committee) and revert the outcomes of such votes if necessary.
Great points - the Security Council is the last resort that needs to step in to halt the outcome of the proposal in the event that it is passed maliciously. That said, such malicious proposals should also be cancellable before the Security Council has to step in with this proposal cancellation function.
Sharing an update after comments from @paulofonseca on the GRC call on proposal cancellation details - sorry I had to clarify most of these with our engineering team and couldn’t answer timely on the call yesterday.
On when the proposal can be cancelled - the proposal can only be canceled during the pending period before voting, and this is enforced at contract level.
On who can cancel - Only the proposal author can cancel proposals, and this is enforced at contract level.
In the event that the proposal author is malicious, and in the edge case that the proposal got majority FOR and achieved quorum, the Security Council would have to intervene to manually cancel the proposal.
Hope this helps! Let us know if there are any additional questions that we have not addressed.
We fully support the approval of this proposal. The ability for the proposer to cancel their own proposal is an essential feature to prevent further issues resulting from mistakes made during on-chain submission.
Although this is not the primary objective of the proposal, it is also important to discuss other ways of enabling proposal cancellation in governance.
As @cliffton.eth pointed out, only the Security Council currently has the authority to cancel a malicious proposal. This creates a reliance on a single protective mechanism within Arbitrum.
Designing redundancies for DAO governance security is essential to maintaining its integrity. This is one of the criteria we consider when evaluating the security of a DAO in our product, Anticapture.
Update: the implementation and security audit for the proposal cancellation work is complete.
The execution payload is currently under a security audit. In accordance to the process laid out in this thread, if DVP Quorum upgrade ( [Constitutional] DVP Quorum for ArbitrumDAO: Implementation & Parameters ) proceeds to an onchain vote, the payloads will be combined, unless a significant portion of delegated voting power (30M ARB) objects by Thursday, Feb. 5th, in which case the proposal cancellation upgrade will be taken through a separate temperature check. This is done to avoid unnecessary voting cycles on an upgrade which has been extensively discussed in the past and the community has reached alignment on the motivation and specification of.
We invite everyone to participate in the upcoming ‘Open Discussion of Proposals Governance Call’ where this upgrade will also be covered.
there’s no need for another temp check for proposal cancellation, since this was already approved, as a temp check, on March 26, 2024, here in this offchain vote: Snapshot
as well as:
Add Flexible Voting to the Arbitrum DAO Governors to enable future innovations like voting from Orbit chains, voting from DeFi contracts, and shielded voting.
Voting in Favor — allowing the proposal submitter to cancel (during the 3-day delegation period) is a strict win, accounting for e.g. human-error cases caught after proposal submission. It also gives the Security Council a cleaner means of canceling proposals if need be.
Point of clarification: the Security Council already has the affordance to cancel proposals; it could be done, e.g., via a proxy-upgrade of the governance contract. This proposal only lets the SC do so in a way that is technically simpler, easier to track after the fact, etc. Thus, the question of under what conditions the SC should be expected to cancel proposals is a perfectly valid one, but one that should be discussed independent of this proposal, as it isn’t relevant
I also support the ultimate decision to carry out this upgrade via a proxy upgrade instead of a redeploy, on the following grounds:
The rest of the governance system uses proxy contracts, and all things equal, we should default to consistency for the sake of minimizing complexity and operational overhead.
Upgrading via redeploying and altering the timelock affordance breaks history tracking, and, more critically, requires the transfer of contract state, which could get especially messy if an update is taking place while other proposals are live.
As far as I can tell, there would be no clear benefit gained from the redeployment route.
Hi everyone, quick update here - as mentioned on the governance call, the proposal cancellation changes have been bundled with the DVP changes on the constitutional onchain AIP that is on Tally at the moment.
The introduction of clearer proposal cancellation authority, particularly where exercisable by the Security Council, appears to formalise discretionary intervention within the governance execution process.
Has consideration been given to how this expanded intervention pathway may affect decentralisation assessments under the EU MiCA framework, especially where a defined body retains structured authority to halt or modify governance outcomes?
While such powers may be justified for security and error mitigation, their existence may be relevant when analysing whether governance remains sufficiently autonomous or is subject to identifiable managerial control.
Meeting Recording