Near Protocol Plans AI Delegates to Represent DAO Members, Could Improve Voter Participation

• AI delegates learn user behavior to automate DAO voting

• Rollout is staged: advisory chatbots → group delegates → one delegate per member

• Participation averages 15–25% in many DAOs; delegates aim to improve representation

AI delegates Near Protocol DAO: learn preferences, vote on behalf of members, reduce voter apathy — read how to prepare your account for delegate options.

What are AI delegates for Near Protocol DAO?

AI delegates are software agents developed for the Near Protocol DAO to represent member preferences in governance votes. They analyze user inputs and historical behavior to recommend or cast votes, aiming to improve low participation and make governance decisions faster and more consistent.

When the AI delegates rollout, it will be done in stages, with early models similar to chatbots, then representing large groups, and finally, each DAO member.

The Near Foundation is developing artificial intelligence-powered delegates to address low voter participation in decentralized autonomous organizations (DAOs). The goal is to turn routine governance decisions into near-instant computations by using agents that know member preferences.

Lane Rettig, a researcher at the Near Foundation specializing in AI and governance, told plain text press that the AI-powered governance overhaul is still in development. The Near Foundation oversees the layer-1 Near Protocol.

How will AI delegates be trained on user behavior?

Delegates are trained by combining explicit user inputs, voting history, and public messages on community channels. This includes interview-style onboarding, past vote records, and contextual signals from social platforms to model political and funding preferences.

Near intends to use a verifiable training approach that provides cryptographic proof of training cycles and inputs to maintain alignment and transparency. Industry data shows adoption of AI agents in crypto has accelerated: investment manager VanEck estimated over 10,000 agents by end-2024 and forecasted significant growth into 2025.

Why will a human remain in the loop?

Near’s researchers emphasize a hybrid approach: AI can handle routine proposals, but humans will retain final authority on critical decisions like fund allocations or major strategy shifts. This preserves accountability and reduces the risk of catastrophic errors from autonomous agents.

Rettig stated that delegates can nudge users and recommend votes, but certain categories of proposals require human judgment to “pull the trigger.” This hybrid model aims to balance efficiency with responsibility.

When will the rollout happen and what stages are planned?

Rollout is planned in stages: early agents will function like chatbots providing context and draft votes, then agents will represent groups with shared preferences, and finally one agent per individual member may be offered. Each stage prioritizes safety, transparency, and verifiable training records.

Near’s main DAO already uses a sentiment and summarization tool called Pulse to surface important content and community trends. Early delegate models will have limited agency and focus on improving information flow and voter engagement before broader autonomy is allowed.

What risks and safeguards are being discussed?

Key risks include misaligned agent behavior, security vulnerabilities, and concentration of influence. Safeguards proposed include verifiable training logs, human-in-the-loop gating for critical votes, and phased deployment with monitoring.

Experts recommend auditability and cryptographic proofs of training data to preserve trust. Platforms and researchers such as OpenAI and industry asset managers have highlighted both the potential and the security challenges of AI agents in decentralized finance.

Frequently Asked Questions

How do AI delegates improve DAO participation?

AI delegates automate routine voting by representing member preferences and nudging informed participation, helping convert low-engagement members into represented votes and increasing effective turnout.

Can delegates vote without user approval?

Near’s stated approach favors human oversight; early phases emphasize advisory roles and explicit consent for automated voting, especially on high-impact proposals.

Are delegate training methods auditable?

The Near team plans verifiable training records that provide cryptographic proof of model inputs and training cycles to maintain alignment and enable audits.

Key Takeaways

• AI delegates: Automate representation by learning member preferences and recommending or casting votes.
• Phased rollout: Starts with advisory tools, proceeds to group delegates, then individual digital twins.
• Human oversight: Critical decisions remain human-controlled; verifiable training and audits are planned.

Conclusion

The Near Foundation’s AI delegate initiative aims to reduce voter apathy and streamline DAO governance by using trained agents that reflect member preferences. With phased deployment, verifiable training, and human oversight, the project balances automation with accountability. Watch for staged releases and governance updates as the system matures.