First, let’s revisit what governance means in DeFi. Governance relies on automatic mechanisms to update the behavior of a DeFi platform in a decentralized way (I’m considering both on-chain and off-chain components, so I refer to a platform rather than just a protocol). This is typically implemented via smart contracts. Users discuss improvement proposals and vote on them according to predefined rules, and the chosen proposals are implemented to achieve the desired new behavior.
But what are backdoors in the context of governance? They are mechanisms that bypass the governance process to change the behavior of a DeFi platform. One might ask: if a governance system is already in place, why use a backdoor? Typically, governance processes take time —time to discuss proposals, create artifacts for each proposal, conduct voting, etc-. However, sometimes changes need to be made more quickly.
For example, what if an exploit is identified that is draining funds from the platform? There may not be time to vote on stopping or fixing the exploit; you need to react immediately. In situations like these, backdoors come into play.
Now, the interesting discussion: we know that urgent situations happen, and we need a fast response. So, why not be prepared for that?
This preface sets the context for answering the following question: Why don’t we rethink our governance to avoid the need for backdoors?
Here are some ideas:
When an urgent situation is detected, the first step is to stop what is causing the problem. This could mean pausing one component, a set of components, or the entire platform. We could start by pausing the entire platform, then gradually restart unaffected parts while leaving the vulnerable component offline until the problem is resolved.
It’s important to consider that fixing issues is not always easy and can sometimes be time-consuming. You could implement temporary fixes to buy more time for a definitive solution.
Do all changes require the same rules for adoption? Is adding a new feature the same as changing a single parameter? Certainly not. Simpler and faster governance rules should be in place. Therefore, governance could have an ‘express mode’ to handle these situations.
We could anticipate suspicious situations and create monitoring systems to detect them. As soon as a monitor detects an issue, it could pause the affected component and raise an alert.
Don’t forget to communicate the urgency to your community, providing constant updates on the progress of clarifying the causes, next steps, resolution, and expected timelines.
Providing your governance with mechanisms for pausing operations, quickly implementing changes, preventing issues, and maintaining good communication will show your community that your platform is prepared for the worst in a world that is always evolving and growing in complexity.
And yes: “Backdoors” are bad.”
