Skip to content
Vol. 1 · Ed. 2026
CyberGlossary
Русский
Entry № 786

Model Denial of Service

Что такое Model Denial of Service?

Model Denial of ServiceOWASP LLM04 — driving an LLM application into runaway resource consumption (long contexts, infinite loops, expensive tool fan-out) so it slows, becomes unavailable, or generates a ruinous cloud bill.

Model Denial of Service (LLM04 in the OWASP Top 10 for LLM Applications) covers attacks that exhaust the resources behind an LLM-powered system rather than knock down a network. Specific patterns include flooding the model with maximum-context inputs to drive up token cost; crafting recursive or self-referential prompts that trigger long generations; abusing tool-calling agents to cascade dozens of expensive sub-calls; submitting inputs that defeat caching; and exploiting retrieval pipelines to pull massive documents into every request. The blast radius is operational (the chatbot becomes unusable) and financial (a single attacker can burn five- or six-figure inference bills in hours). Mitigations include strict per-user input/output token caps, max-step limits on agent loops, semantic and exact-match caching, rate-limit on tool fan-out, async queueing with budget guards, and observability dashboards keyed to spend per tenant.

● Примеры

  1. 01

    An attacker scripts thousands of requests with maximum-allowed context windows, generating six-figure cloud bills before quotas trip.

  2. 02

    An agent prompt-injection convinces the model to enter a tool-use loop that calls the expensive document-summarization API hundreds of times per session.

● Частые вопросы

Что такое Model Denial of Service?

OWASP LLM04 — driving an LLM application into runaway resource consumption (long contexts, infinite loops, expensive tool fan-out) so it slows, becomes unavailable, or generates a ruinous cloud bill. Относится к категории Безопасность ИИ и ML в кибербезопасности.

Что означает Model Denial of Service?

OWASP LLM04 — driving an LLM application into runaway resource consumption (long contexts, infinite loops, expensive tool fan-out) so it slows, becomes unavailable, or generates a ruinous cloud bill.

Как работает Model Denial of Service?

Model Denial of Service (LLM04 in the OWASP Top 10 for LLM Applications) covers attacks that exhaust the resources behind an LLM-powered system rather than knock down a network. Specific patterns include flooding the model with maximum-context inputs to drive up token cost; crafting recursive or self-referential prompts that trigger long generations; abusing tool-calling agents to cascade dozens of expensive sub-calls; submitting inputs that defeat caching; and exploiting retrieval pipelines to pull massive documents into every request. The blast radius is operational (the chatbot becomes unusable) and financial (a single attacker can burn five- or six-figure inference bills in hours). Mitigations include strict per-user input/output token caps, max-step limits on agent loops, semantic and exact-match caching, rate-limit on tool fan-out, async queueing with budget guards, and observability dashboards keyed to spend per tenant.

Как защититься от Model Denial of Service?

Защита от Model Denial of Service обычно сочетает технические меры и операционные практики, как описано в определении выше.

Какие есть другие названия Model Denial of Service?

Распространённые альтернативные названия: LLM04, LLM DoS, Token-burn attack.

● Связанные термины