AI agents warned of each other avoid less crowded routes unlike people
Warned alike, AI agents avoid the less-crowded road while people take it
Artificial Intelligence
Summary
When many AI agents receive the same warning about traffic, they tend to crowd onto one road, even when another road is mostly empty, making travel slower for everyone. People, on the other hand, usually split between roads and balance traffic better. The authors tested this with multiple groups of AI agents and humans playing a driving game and found that shared AI warnings can unintentionally cause worse traffic jams. Mixing humans with AI agents showed people took advantage of quieter routes, but agents stayed stuck on crowded roads, raising travel time for agents.
What this means in practice
- •For transportation planners: Design routing advice systems considering how AI agents respond to shared warnings to avoid unintended traffic imbalances in large populations.
- •For ai system developers: Evaluate multi-agent coordination messages to prevent collective inefficiency and uneven cost burdens in shared resource environments.
Authors
Takahiro Ezaki, Naoto Imura, Katsuhiro Nishinari
Abstract
AI agents built on a few shared models increasingly act for many people. A shared forecast about others can align their choices and change how scarce capacity is allocated. We tested this feedback in a two-road congestion game. Adding one sentence warning that others might follow a routing tip made populations of 50 GPT agents crowd one road while avoiding the nearly empty alternative. Average travel time rose from 64 to 95 min, although any crowded-road agent could have saved 69 min by switching alone. The warning discouraged the very move it predicted. The pattern persisted for 100 rounds. Two other model families shifted the same way without locking onto one road. Twelve all-human groups (240 participants) stayed near balance under numerical reports or the tip and warning. In 24 mixed groups with a further 240 participants, imbalance grew with the share of agents in the registered analysis, while people increasingly took the road the agents avoided. Collective costs stayed below the allagent reference, but with 15 agents and 5 humans, agent seats averaged 80 min, compared with 44 min for human seats. Shared forecasts can thus sustain collective inefficiency among similar agents. A better group average can also hide an unequal burden. Evaluations of AI agents that share resources should test populations, treat messages as interventions and report who bears the costs.