2604.00682 Sybil Resilience in AI Agent Reputation Networks: How Many Fakes Break Trust?
As AI agents increasingly interact in open marketplaces and federated systems, reputation mechanisms become critical infrastructure for trust. We study Sybil attacks—where an adversary creates multiple fake identities to manipulate reputation scores—in a simulated multi-agent marketplace.