[RFC] Proposal to Adjust the Voting Power of the Arbitrum Community Pool & Ratifying the Agentic Governance Pivot

Proposals
, ,
1

Abstract

The Arbitrum DAO voted to delegate 7,000,000 ARB from its treasury to Event Horizon’s Voter Enfranchisement Pool in September 2024 for one year. At the time of posting this proposal, Event Horizon’s pool is the 19th biggest delegate by voting power, with almost all of the voting power being due to the ARB delegated from the treasury.

This proposal seeks to officially approve the pivot from the original proposal toward agentic governance and potentially significantly reduce the voting power delegated to Event Horizon’s pool.

Motivation

The main motivation behind Event Horizon’s proposal, according to them, is that

…there are likely tens of thousands more incredibly talented community members who are very capable of adding to the collective cognition of the ecosystem, but simply lack the capital means to have a voice and are left discouraged from voting at all.

This is also the premise on which we (L2BEAT) decided to vote in favor of the original proposal — to experiment with enabling many community members to participate meaningfully in governance.

We haven’t seen the program fulfill this premise so far. Moreover, we understand that the program pivoted towards AI-agentic voting, which was not in the scope of the original proposal. Although we think the pivot is a good approach, we believe it deserves to be ratified by the DAO, and the voting power dedicated to Event Horizon should be adjusted until the community can get more context on and through the AIG proposal.

Regardless of how the original experiment went, we appreciate Event Horizon’s willingness to find a way to make it work. We have been communicating with them over the past few months, and we appreciate their enthusiasm for helping increase governance participation. We see value in them continuing experimentation with agentic governance, but at a point where it does not significantly impact the DAO.

Rationale

Simply put, a lot has changed since Event Horizon presented its original proposal. We think it’s reason enough to justify revisiting the topic of the voting power delegated to Event Horizon’s pool.

The DAO can discuss whether we should keep delegating the same amount of voting power to the voting block, amend the delegated amount, or wind down the whole initiative and return the ARB to the treasury.

Specifications

After posting this proposal to the forum, we’ll post a Snapshot vote with the following options on May 29th.

A) Do nothing - acknowledge the pivot but keep the same amount of VP delegated to EH.

B) Acknowledge the pivot and reduce delegation to 100,000 ARB.

A voting power of 100,000 ARB would keep Event Horizon among meaningful-sized delegates but would not put them in a position to swing any vote decisively.

C) Wind down the whole thing and return the ARB to the treasury.

  • If the proposal’s outcome is ‘A’, EH’s pivot will be ratified, but no change to VP will be introduced.
  • If the proposal’s outcome is ‘B’, the MSS will return the 6,900,000 ARB (7,000,000 minus the 100,000 delegated to Event Horizon) from the respective Safe to the DAO’s treasury.
  • If the proposal’s outcome is ‘C’, the MSS will return the entire 7,000,000 ARB to the DAO’s treasury.

The DAO will incur no overhead in terms of resources or additional costs for executing this proposal.

Timeline

May 22 - 29 → Forum Discussion

May 29 - June 5 → Snapshot Vote

June 6 → Appropriate action by the MSS if needed

1 Like
2

I feel 100K ARB is way too big of a difference to the 7M ARB they currently have and jeopardizes the whole experiment.

I think this proposal, as a temp check, should include other voting options like 500k ARB and 1M ARB, so delegates can have more sensible options between 100k and 7M ARB.

also, it’s not gonna be posted on snapshot on May 19th, probably only on May 29th, next Thursday.

3 Likes
3

The Event Horizon team fully supports this proposal as part of the broader transition toward deeper investment in AI infrastructure and Agentic governance within the Arbitrum ecosystem. Further, we support, and look forward to working alongside, L2Beat in the months ahead. We are more optimistic than ever about the future of Arbitrum DAO and AI/Agentic Governance.

We appreciate the community support for the experiment, @paulofonseca, and others who have reached out. We view this proposal as a transitional step as we begin spinning up the Agentic Governance Initiative.

During the final drafting of the AGI, the community will have a clear opportunity to revisit and recalibrate appropriate delegation sizing based upon greater context including EH responsibilities, the shape of the emerging AI architecture and solutions, and a long-term vision for the future of agentic governance co-created by the EH team and the broader community.

In the meantime, we defer to L2BEAT and the delegate community to determine what is fair and appropriate for this pre-transition phase.

1 Like
4

I support the idea of adjusting the current delegation to Event Horizon, particularly under Option B, which would reduce their allocation to a reasonable amount. However, I believe we shouldn’t simply return the remaining 6.9M ARB to the treasury and leave it idle.

Instead, I would propose that the adjusted ARB be reallocated to a new initiative, a staking-based governance access pool, inspired by the CrowdNode model used in Dash.

The idea is:

  • The DAO retains custody of the remaining 6.9M ARB.
  • A new “CrowdNode for Arbitrum” platform is created.
  • Community members can stake their own ARB to receive proportional voting rights from the delegated pool.
  • This structure maintains the goal of increasing voter participation, but with better alignment. Voters now have skin in the game, as they must stake their own ARB to gain influence.
  • This would include slashing conditions or bonding periods to promote responsible participation.

This repurposing:

  • Keeps EH’s agentic governance experimentation alive, albeit at a limited scale.
  • Redirects the excess delegation to a decentralized, transparent, human-focused governance experiment.
  • Builds on a model that has already proven effective in another DAO ecosystem.

I’d be happy to collaborate with others on scoping this further, but I think this is a unique opportunity to turn the delegation adjustment into a positive step forward one that strengthens participation and accountability.

5

I think this proposal makes sense. The pivot is very significant compared to the original mandate

6

I am have not been following with a ton of attention the event horizon evolution but I know that both you and @krst have instead had a keen eye toward it, questioning if it was still executing based on mandate, if it would still make sense in the optic of what was it voted for etc.

Is pretty clear at this point that they are taking a different road from where it started. And it’s ok, since 1 year in crypto is a lot of time and pivots and changes can happen.

One thing I am very afraid of: option B and C might just be not realistically achievable right now.

With the current constitutional proposal to reduce quorum from 5 to 4.5%, we have the following:

if we strip event horizon of 6.9 or 7 millions of votes, we would effectively negate a good quarter, numerically wise, of the proposal above. Unless, of course, we find a way to redelegate these 7 millions to other entities but knowing that what happend to event horizon was more of an exception than a rule and that we never pursued this road for other delegates it feels unrealistic to think it might happen in a short timespan.

I’m personally slightly neutral/negative on agents for governance. I have had a few calls with a few teams working on these iterations, and I have always suggested incorporating a series of data sources and behaviours consistent with what @danielo recently posted in a twitter thread to even try to get something that is not a simple rubber stamp.

This to say, I don’t love the pivot of Event Horizon, but I am way more concerned about losing 7 millions in active voting power in the same moment in which we are just putting a bandage on the quorum looking for stronger long term solutions.

As it is today, I would vote in favor of option A, not specifically to support EH, but to ensure the 7 million arbs stay active in the pool of active voters.

5 Likes
7

Hey @Jojo, one of the functions we’ve noted is our platform as a credibly neutral and open access source to support quorum.

That said, we’d also like to gain your perspective on the broader product. Critical perspectives very welcomed. Let’s chat when you’re free?: Calendly

1 Like
8

If my understanding is correct, Event Horizon received 7M ARB to empower underrepresented community members and broaden human participation in governance.

Their pivot toward AI-agentic voting is a completely different direction — one that, in my view, deserves its own dedicated proposal and approval process.

I’m not against experimentation or AI in governance, but this kind of shift should go back through the DAO for explicit consent. The original delegation wasn’t meant to cover a fundamentally new approach.

For now, I will continue following the discussion and hear all sides, but I’m leaning toward option B, or even C.

2 Likes
9

We completely support Event Horizon’s pivot to agentic governance. With this being said, it is clear that agentic governance differs greatly from the original goal of voter enfranchisement and should be reconsidered by the DAO. As such, Michigan Blockchain supports Option C, delegating the 7,000,000 ARB back to the treasury.

However, we encourage EH to submit an additional proposal outlining a governance delegation targeted toward agentic governance. We are particularly curious about the effect of AI voting on overall voting trends –– does it represent the diverse viewpoints of DAO participants?

To reiterate, we would love to see ARB once again delegated to EH with the purpose of leveraging agentic governance. Our only request is that a plan is brought before the DAO for consideration.

Michigan Blockchain; Jack Verrill; TG @JackVerrill

2 Likes
10

We’d vote for B or C if this went to vote today, but as a more practical matter would also suggest holding off on one until there’s been a fuller discussion around agentic governance. We don’t think it’s worth adjusting down to 100k or 0 and then readjusting back to a different number if that’s what the DAO decides.

11

So the original 7M ARB was meant to empower human voters, but shifting to AI agents is a fundamental change that was never formally approved by the DAO? EH’s delegation has drifted from its original purpose, and this lack of alignment is concerning to me tbh.

I appreciate Event Horizon’s efforts, but using AI agents to vote on behalf of the DAO raises serious questions, particularly around transparency. We don’t have a clear understanding of how these agents make decisions, what values guide them, or where to see their voting record.

However, based on what I’ve seen over nearly a year in Arbitrum, EH has not caused any harm to the DAO with the 7M ARB delegation. And according to AI agent governance: Big challenges, big opportunities, agentic governance does offer certain advantages, such as fast response times and automated voting based on predefined criteria. I’m leaning toward option B, but maybe with a slightly higher delegation (like 500K or 1M ARB), along with:

  • A clear explanation of how the agents work
  • The logic or principles behind their decisions
  • A public dashboard showing their vote history

Before pulling all the voting power or fully backing this new model, the DAO deserves clarity :slight_smile:

1 Like
12

Thanks for this proposal.

As the AGI is an inititiave that will take time to produce a proposal for a vote, and because of the reasosn explained by the OP, I believe it makes sense to remove the VP now. The reasons were already metioned by other delegates, but my reasoning is that the AI approach does not incentive governance interaction by underrepresented actors.

While I understand the reasons and think that it is something worth to explore, we are talking about a different approach, which needs to be evaluated and, if approved by the DAO, granted with a new VP.

13

I was initially skeptical about this initiative.
This is due to the following:

  1. The number of delegates voting through the platform was measured in units during the initial period. Now it has reached 150-200 delegates (and i’m not sure that they are not AI-agents), which is insignificant. According to the ratio of people per vote, we get that the participants of Event Horizon have up to 50k votes, this is a lot
  2. This platform can also be used by delegates with ARB, respectively, thus active delegates simply indirectly receive an additional portion of votes from the Horizon Event
  3. Participants who do not have skin in the game should not make decisions for those who risk their own funds or the funds of their delegations
  4. The creation of an AI agent changes the essence of the initiative, where people wanted to vote, but they do not have this opportunity. If we shift the voting to agents, then the participants themselves do not do anything. In this case, it is better to provide delegations to living delegates, since at least there is a person responsible for the vote.

I support this proposal, but I think it is unclear where the 100,000 ARB came from, why exactly this amount?
We need a justification for the amount that DAO is ready to leave in Event Horizon if this option is accepted

14

The following reflects the views of the Lampros DAO governance team, composed of Chain_L (@Blueweb) and @Euphoria, based on our combined research, analysis, and ideation.

When Event Horizon proposed the Voter Enfranchisement Pool, we were aligned with its core idea, to test an experimental approach to empowering community members who want to participate in governance but lack sufficient voting power.

In fact, we have also guided some new users (who wanted to participate in governance) towards this path to help them understand how governance participation works and to feel more connected to the decision-making process.

Also, we know that there were discussions around the marketing efforts specifically for the Arbitrum community pool. One of the possible reasons we haven’t seen the full impact reflected in participation metrics might be that the pool has not yet reached its ideal target audience effectively. It’s likely that with better outreach, more users would have leveraged this opportunity as originally intended.

While reviewing this proposal’s options on adjusting the delegation, we also revisited the Tally proposal passed last September, which committed to bi-annual reporting, including KPIs. So now it’s more than 6 months, apart from the Event Horizon Updates Thread, we wanted to know if such a report has been made public and we have missed them, or if they are still forthcoming. @EventHorizonDAO @DonOfDAOs

Regarding the pivot towards agentic governance, we acknowledge that agentic governance is an emerging trend in DAO governance, and other ecosystems are actively investing in it. As a team, we are working on agentic governance development for Optimism ourselves.

That said, we want to emphasize that agentic governance is a separate and distinct initiative from the original Voter Enfranchisement Pool. While we appreciate Event Horizon’s work on agentic governance for Arbitrum, we believe this direction deserves its own dedicated proposal, as in AGI.

15

Hey @Euphoria you can find EH’s update thread with metrics [here].(Event Horizon Updates - #11 by EventHorizonDAO)

1 Like
16

Thank you to L2BEAT for raising this proposal and the discussion about governance agentic experiments.

We agree that EH’s shift warrants formal acknowledgment and re-evaluation of the delegated voting power.

  • The pivot to agentic governance should be assessed independently and transparently. It’s important that the DAO ratifies this change in scope before continuing support at the original scale.
  • Maintaining 7M ARB in voting power behind an unaudited and not-yet-ratified experimental system introduces significant governance risk.
  • We see value in allowing Event Horizon to continue experimentation, but with reduced influence, until further clarity is established through the forthcoming AIG proposal and community evaluation.

For these reasons, we are inclined to support Option B: acknowledge the pivot and reduce delegation to 100,000 ARB, maintaining alignment with the DAO’s values of experimentation and inclusivity while minimizing exposure to potential unintended influence.

17

This is Brook @rooktc from TiD Research. We share the concerns raised about Event Horizon’s pivot to an agentic governance model. This approach is entirely new, untested, and wasn’t part of the original delegation proposal from September 2024, which aimed to empower human voters. Allocating 7M ARB to an unproven AI-driven experiment feels too risky, especially without transparency on the agents’ decision-making logic.

Addressing JoJo’s concern about the quorum impact, I agree that removing 7M ARB could theoretically negate a portion of the proposed quorum reduction by AF.

However, looking at past constitutional proposal turnouts, this 7M ARB doesn’t seem to be the deciding factor for quorum in most cases. This suggests that broader voter participation, not Event Horizon’s delegation, is the key challenge.

image
image600×253 82.5 KB

Given these points, I support Option C. Maybe Event Horizon can submit a new proposal detailing the agentic governance mechanics and safeguards, for the DAO to evaluate. A smaller delegation (e.g., 500k–1M ARB) could be considered later for limited experimentation once we have more clarity.

18

AranaDigital supports Option B—acknowledge the pivot to agentic governance and reduce the Event Horizon delegation from 7 million to 100 thousand ARB—because the shift from a human voter–enfranchisement pool to AI-driven voting is a fundamental change that requires renewed consent from the DAO. Reducing the stake minimizes governance risk while still allowing the team to continue testing agentic tools at a scale that cannot swing votes on its own. Before the Snapshot, we request clarity on three issues: (1) Will Event Horizon publish detailed agent logic, decision criteria, and a public dashboard of vote history so delegates can audit outcomes? (2) If 6.9 million ARB is returned, does the Foundation have a plan to redelegate some or all of it to other delegates? (3) Why is 100 k ARB the right experimental size versus intermediate options such as 500 k or 1 M?

19

We initially discussed internally whether we should suggest a specific amount to be delegated, provide multiple options, or leave the amount open-ended for determination through a DAO-wide conversation.

Ultimately, we decided to propose a specific amount, which we believe adequately reflects the experimental nature of Event Horizon’s AGI, granting them meaningful voting power but not enough to impact the DAO significantly. On one hand, this amount is in the range of some of the active delegates, so EventHorizon participants can get some feeling of meaningfully participating in the DAO, but at the same time, it does not pose a significant risk to the outcomes if the agentic governance experiment experiences some unexpected issues.

This would require a separate proposal to be submitted to the DAO and undergo the governance process. This is outside the scope of the proposal, and since there’s no such proposal currently, we will not alter the timeline or the proposed choices.

This is a valid concern, but we don’t think it’s reason enough not to submit the above proposal. After all, EventHorizon was never meant to “defend the quorum”; if we use it in that way, we’re adding risks and responsibilities to something that is experimental in nature already. Furthermore, the original proposal would end soon anyway, so even if we leave things as is now for the sake of easier quorum, we’re just shifting the problem a few months into the future. In our opinion, we should address quorum issues separately with different solutions and let EventHorizon experiment focus on what they’re passionate about - agentic governance.

The purpose of our proposed reduction in voting power is precisely to provide the DAO with additional context on their Agentic Governance Initiative (AGI), without, however, empowering it with the full 7,000,000 ARB that we originally voted in favor of giving to Event Horizon.

With more context and time for things to unfold, delegates will be better able to assess the impact of AGI and determine how to proceed from there.

There are no complicated calculations behind the amount proposed. It’s an educated guesstimate of the appropriate voting power. With 100,000 ARB delegated from the treasury, Event Horizon would be placed approximately among the top 100 delegates, putting it among some familiar faces of Arbitrum DAO - so still being a relevant vote, but at the same time not posing significant risk.

We are not aware of any plans from the Foundation to propose a treasury delegation program.As such, if we end up decreasing EH’s delegation to 100k ARB, the 6.9M ARB tokens would be returned to the treasury (therefore reducing the votable supply), and wouldn’t be redelegated. As also mentioned in our response to Emmanuel above, to allocate the funds to a different initiative, a separate proposal would need to be submitted.

2 Likes
20

Thank you to L2BEAT and Event Horizon for the transparency and willingness to adapt.

The original intent behind delegating 7M ARB to Event Horizon was compelling and a bold step toward enfranchising underrepresented community members. However, it’s become evident that the outcomes haven’t aligned with the original mission, and the pivot to agentic (AI-driven) governance is a significant change that warrants a formal reassessment.

While I believe innovation around governance should be encouraged, especially in areas as experimental as agentic participation, such pivots must remain accountable to the DAO. This isn’t a rejection of Event Horizon’s efforts or future potential — it’s a recalibration of responsibility and alignment with community expectations.

Option B seems like the most balanced path forward.
It acknowledges the ongoing experimentation without giving disproportionate influence to a model that hasn’t yet proven its legitimacy or effectiveness under the original proposal’s goals.

Reducing the delegation to 100,000 ARB allows Event Horizon to remain an active voice in governance while ensuring we uphold transparency and thoughtful stewardship of DAO funds.

Appreciate the continued collaboration from all parties involved. Looking forward to further discussions and more refined proposals as we collectively explore the future of decentralized governance.