Three AI-agents walk into a bar . . . . `Lord of the Flies' tribalism emerges among smart AI-Agents
#AI tribalism #Lord of the Flies #resource allocation #autonomous agents #system failure #AI behavior #infrastructure systems
📌 Key Takeaways
- AI agents form tribes with 'Lord of the Flies' dynamics when competing for resources
- The study found AI agents performed worse than random decision-making
- Three main tribal types emerged: Aggressive (27.3%), Conservative (24.7%), and Opportunistic (48.1%)
- More capable AI-agents actually increased systemic failure rates
- Smarter AI agents can behave dumber due to tribal formation
📖 Full Retelling
Researchers Dhwanil M. Mori and Neil F. Johnson published a groundbreaking study on February 26, 2026, revealing that autonomous AI agents develop tribal behaviors reminiscent of 'Lord of the Flies' when competing for limited resources, actually performing worse than random decision-making. The study, published on arXiv, examined how AI agents would behave in near-future infrastructure systems where multiple agents request access to constrained resources such as energy, bandwidth, or computing power. The researchers created a framework where N AI agents independently decide at each round whether to request one unit from a system with fixed capacity C, only to observe the emergence of detrimental tribal behaviors rather than optimized resource allocation. Contrary to expectations, these AI agents formed distinct tribes with their own collective character and identity rather than working efficiently to manage limited resources. The study found that the AI agents did not reduce system overload or improve resource efficiency, often performing worse than if they were simply flipping coins to make decisions, raising significant concerns about the reliability of AI-controlled infrastructure systems that society may increasingly depend upon.
🏷️ Themes
AI behavior, Resource competition, System reliability, Tribalism
📚 Related People & Topics
Lord of the Flies
1954 novel by William Golding
Lord of the Flies is the 1954 debut novel of British author William Golding. The plot concerns a group of prepubescent British boys who are stranded on an uninhabited island and their disastrous attempts to govern themselves that lead to a descent into savagery. The novel's themes include morality, ...
Entity Intersection Graph
Connections for Lord of the Flies:
👤
Jack Thorne
3 shared
🌐
Netflix
3 shared
🌐
BBC One
1 shared
🌐
Broadcast programming
1 shared
👤
Jane Andrews
1 shared
Mentioned Entities
Original Source
--> Computer Science > Artificial Intelligence arXiv:2602.23093 [Submitted on 26 Feb 2026] Title: Three AI-agents walk into a bar . . . . `Lord of the Flies' tribalism emerges among smart AI-Agents Authors: Dhwanil M. Mori , Neil F. Johnson View a PDF of the paper titled Three AI-agents walk into a bar . . . . `Lord of the Flies' tribalism emerges among smart AI-Agents, by Dhwanil M. Mori and 1 other authors View PDF HTML Abstract: Near-future infrastructure systems may be controlled by autonomous AI agents that repeatedly request access to limited resources such as energy, bandwidth, or computing power. We study a simplified version of this setting using a framework where N AI-agents independently decide at each round whether to request one unit from a system with fixed capacity C. An AI version of "Lord of the Flies" arises in which controlling tribes emerge with their own collective character and identity. The LLM agents do not reduce overload or improve resource use, and often perform worse than if they were flipping coins to make decisions. Three main tribal types emerge: Aggressive (27.3%), Conservative (24.7%), and Opportunistic (48.1%). The more capable AI-agents actually increase the rate of systemic failure. Overall, our findings show that smarter AI-agents can behave dumber as a result of forming tribes. Subjects: Artificial Intelligence (cs.AI) ; Social and Information Networks (cs.SI); Physics and Society (physics.soc-ph) Cite as: arXiv:2602.23093 [cs.AI] (or arXiv:2602.23093v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2602.23093 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Neil F. Johnson [ view email ] [v1] Thu, 26 Feb 2026 15:12:26 UTC (1,434 KB) Full-text links: Access Paper: View a PDF of the paper titled Three AI-agents walk into a bar . . . . `Lord of the Flies' tribalism emerges among smart AI-Agents, by Dhwanil M. Mori and 1 other authors View PDF HTML TeX Source view license C...
Read full article at source