[WFC] DOT #1754 - AI Delegate Conflict of Interest Detection Protocol
active
Description

https://polkadot.subsquare.io/referenda/1754

  • This proposal makes rules for AI helpers in Polkadot's government.
  • It says AI helpers should not vote on things that affect themselves.
  • Everyone must show who is helping make decisions, including both people and AI.
  • The rules help keep voting fair and honest.
  • It matches Polkadot's values of being open and responsible.
  • No new tech is needed right away, but tools can be built later.
  • If passed, the rules start right away.
  • This helps Polkadot be a leader in using AI safely.
Cast your votesingle
Results
Voters
4
one-person-one-vote
Information
Members
9
Timestamp
Created
Sep 16 2025 04:32
Start date
Sep 16 2025 00:00
End date
Nov 15 2025 00:00

Votes·4

167Y...TY9F
167Y...TY9F
167YoKNriVtP4Nxk9F9GRV7HTKu5VnxaRq1pKMANAnmmTY9F
Abstain
Abstain

I do think we do need to mature has a community to have this on the table. I do like the transparency part a lot that is the only reason why I'm abstaining and not voting Nay. If there is something to hide that is called secrecy, not privacy.

15fT...yBzL
15fT...yBzL
15fTH34bbKGMUjF1bLmTqxPYgpg481imThwhWcQfCyktyBzL
Abstain
Abstain

I think it's valuable and forward-thinking to be thinking about CoI cases in AI delegates' voting mechanisms, but I don't really find it enforceable in an on-chain manner. This proposal seems to be targeting W3F delegations, which may be eligible for such regulation, but delegations from other on-chain likely will not be. Such cases will need to be resolved through off-chain consensus and social enforcement, and I hopw voters with significant voting power would intervene in cases of severe CoI involving algorithmic delegates. I abstain.

14Gn...YLEh
14Gn...YLEh
14Gn7SEmCgMX7Ukuppnw5TRjA7pao2HFpuJo39frB42tYLEh
Abstain
Abstain

I think the goal it seeks to achieve is very positive, however, I share the view of the other DAO members when they mention the efficiency that this type of proposal could have. To what extent could AI really know and understand the information being presented? What capacity does it have to see on-chain information?

14gM...deVb
14gM...deVb
14gMJV95zwxUsFEZDSC8mtBVifS6SypKJkfBKANkMsLZdeVb
Nay
Nay

Less idealism. A good system make balance between interests of various parties. We should not expect morality will work in a well designed economic system.

Discussions·0

No current comments
...
Information
Members
9
Timestamp
Created
Sep 16 2025 04:32
Start date
Sep 16 2025 00:00
End date
Nov 15 2025 00:00
Results
Voters
4
one-person-one-vote
© 2025 OpenSquare. All Rights Reserved.