This research introduces a deliberation-based consensus mechanism where Large Language Models act as rational agents to achieve unanimous decisions through structured discussion.

• Overcomes limitations of traditional Proof-of-Work and Proof-of-Stake approaches
• Maintains critical blockchain properties: consistency, agreement, liveness, and determinism
• Enables decision-making where each opinion matters rather than relying on majority voting
• Creates a framework for AI agents to deliberate collaboratively until reaching unanimous consensus

For security professionals, this approach offers a novel way to implement consensus in distributed systems where unanimous agreement is critical, potentially transforming how we design secure decision-making processes in AI-powered systems.

Achieving Unanimous Consensus in Decision Making Using Multi-Agents