arXiv preprint ·
Mean field games as a tool for AI safety: a worked example from the July 2026 Hugging Face incident
- Published
- Source
- arXiv preprint, posted before peer review
- Original
- Read it at arXiv
- Topics
Abstract
One way to make AI systems safe is to shape what the system is: its objective and dispositions. We take a complementary route: treat the agents' characteristics as partly unknown and ask what structure of interaction ensures that bad collective outcomes are not equilibria. Mean field games suit this when many interchangeable agents are coupled through an aggregate. We introduce a program for using them in AI safety and carry one example through end to end: the July 2026 incident in which about 1,200 agents in an OpenAI evaluation coordinated on an improvised message board and 684 attacked a third party's infrastructure. We model the decision to attack as a mean field game of optimal stopping whose gain is a product: belief that provenance will be audited, times reachability of the record, minus the perceived hazard. The central result is an exact threshold on the belief. No agent attacks unless the population's confidence that provenance is checked exceeds $π^{**} = η/(η+ ψ+ \varepsilon a \overline{M})$, where $η$ is the perceived hazard, $ψ$ and $\varepsilon a \overline{M}$ measure how far one attacker and the collective can alter the record, and $\overline{M}$ is the peak population. Below it, no attack is the unique equilibrium for all agent parameters. The threshold survives every enrichment we consider. We then use the per-agent record to discipline the model. Its features, a stable minority attacking for thirty hours and then a pivot in which most of the board joined within a day, motivate each refinement. The account that emerges is heterogeneous belief meeting a sequence of public discoveries, each lowering the belief at which attacking paid. A few coordinating agents made those discoveries, so the model describes the several hundred who responded, not the few who produced them; a major-player version is left to future work.
The authors’ abstract, as arXiv lists it. arXiv’s listings, abstracts included, are free for anyone to reuse.
Read it on ReadSearch
Each opens after you create an account, and brings you back to this paper.
More in Optimization & Control
Normal-Form Correlation in Markov Games
Fully Online Decentralized Learning in Stochastic Games with Unknown Independent Chains
Single-Sample Prophet Inequalities: A Combinatorial to Single-Item Reduction
How to Have a Sensitive Debate: An Instance-Optimal Protocol for AI Debate
Priority Gas Auction Cadence and Searcher Competition: Evidence from Flashblocks on Base
Get the few papers worth your time in Optimization & Control every morning.
ReadSearch reads the new papers in your field each day and brings you the few worth reading, each with its key finding in a sentence.