Fault Tolerance and Liability Apportionment - Akash Narayan

ABSTRACTWhen a chain of agents and APIs produces a loss, two bodies of expertise arrive and talk past each other. Reliabilityengineering treats the event as a fault to be contained, and reaches for retries, circuit breakers and compensatingtransactions. Law treats it as a question of attribution, and reaches for defect, negligence and the allocation of providerstatus along a value chain. This paper argues that both under-perform on agent chains for one shared reason: their centralassumptions were built for components that fail visibly and stop. An agent that has gone wrong does not stop. It returnsoutput that is well formed, confident, plausible and incorrect, which is closer to a Byzantine fault than a crash, and neitherthe engineering toolkit nor the doctrine of defect has an affordable answer to Byzantine faults. From this the paper drawsa claim about where apportionment will actually happen. The cost of reconstructing causal share rises with the number ofhandoff pairs while the value of a typical individual loss does not, so beyond a shallow depth after-the-eventapportionment costs more than the loss it resolves. Liability will therefore be allocated in advance, by protocol andcontract, in the manner the card networks used for counterfeit fraud, and allocation rules will attach tomachine-observable facts rather than to litigated findings. The paper sets out what a chain must emit for such rules tobind, argues that bounded and attributable loss is a more useful design objective than low error rates, and treats the sharedmodel substrate beneath nominally independent parties as the systemic risk that the insurance market has already begun toprice by withdrawal. Keywords: multi-agent systems; liability apportionment; fault tolerance; Byzantine faults; agentic AI; product liability; systemic risk;blast radius; attribution; API value chains Author: Akash Narayan Disclosure by Author: Portions of this manuscript were prepared with the assistance of generative AI tools for research synthesis, drafting, and editing. The models used were Indian Sovereign AI models provided by Ayen (https://ayen.in/).