Explore 9 source-supported AI security cases for LLM06:2026 · Unbounded Consumption in OWASP LLM Top 10 (2026), with evidence, scope, and defensive analysis.
Risk category
Unbounded consumption occurs when an LLM application does not adequately limit resource use, enabling excessive inference, unsustainable charges, service disruption, or model cloning. A low-cost request can trigger disproportionately expensive token generation, extended reasoning, multimodal processing, or a sequence of downstream tool and agent operations. Shared infrastructure and cloud billing can spread the operational or financial impact beyond the original request. The risk therefore depends on resource and cost boundaries across the full workflow, not only the number of incoming requests.
Google Threat Intelligence Group and Mandiant reported LLMJacking across several 2026 investigations: actors used other organizations' credentials to run cloud language models at the victims' expense. Some incidents incurred more than US$100,000 in a day. A separate token-routing service handled over 500,000 calls in less than eight hours, passing nearly US$1 million in costs to one victim. The investigations show several routes into cloud or AI gateways and different ways to cash out. For LiteLLM operators, v1.83.14-stable includes the fix for the cited CVE but still lacks the key ownership-change check added in v1.84.0.
Why it relates: GTIG documented unauthorized consumption at scale. The Jet Night case alone incurred about US$1 million after over 500,000 requests in less than eight hours.
Evaluate each observed pattern with its own supporting investigation. The reported key-regeneration behavior and the case-wide advisory identifier are distinct source facts.
Railway is an application deployment platform. Railway reports an October 20 partial build-host filesystem escape, misuse of a build credential against Temporal and theft of its internal monorepo. After that route was contained, an erroneous OIDC binding enabled a second entry to production credentials and variable decryption. At most 1.85% of workspaces could have had variables decrypted; this is an upper bound, not proof that every variable was stolen. Subsequent misuse of AI keys was observed.
Why it relates: Stolen AI credentials were used outside Railway and generated substantial unauthorized AI-service charges. The entry was a conventional hosting-platform breach, not an LLM prompt exploit. The reported charge is an estimate, and later key users are not all the original intruder.
AI relevance comes from exposed credentials and subsequent AI-service abuse. The original platform entry is not an LLM exploit.
Physician and developer Lin reported about NT$30,000 in unauthorized Vertex AI charges within an hour across his Google Cloud projects. His first-person account, republished by Business Next, links those calls to unfamiliar service accounts (machine identities) created roughly two weeks earlier; Google subsequently adjusted most charges. The case concerns cloud AI resource theft, while the initial compromise path and any attacker use of AI agents remain unverified.
Why it relates: The account holder reports unauthorized Gemini Flash and video-generation inference charges of about NT$30,000 within an hour. Paid model consumption continuing despite an alerts-only budget is a direct financial resource-consumption outcome; no service outage, optimized sponge input or agent loop is inferred.
A source-reported Google Cloud AI resource-theft incident. Mappings cover paid model consumption, unauthorized cloud administration and inference access. The MAESTRO overlap is limited to infrastructure resource hijacking; it does not assert an agent runtime.
NSA, CISA and FBI attributed coordinated distillation campaigns to six China-based AI companies. Their September 2026 advisory describes fraudulent accounts, bulk subscriptions, gray-market proxies, cross-provider routing, request-metadata removal and hidden-reasoning prompts, followed by response collection and quality checks. Distillation trains a model from another model’s outputs; the advisory’s attribution and capability-assessment claims are attributed to the agencies, not presented as independently reproduced measurements or model-weight theft.
Why it relates: the campaigns attributed by NSA, CISA and FBI distributed repeated inference across account routes to acquire outputs for model cloning.
the campaigns attributed by NSA, CISA and FBI; capability distillation is distinguished from stealing original weights. Attribution and capability outcomes remain agency assessments. Hidden-reasoning prompts are attempts, not proof that the returned text was an internal trace. No autonomous agent deployment is inferred.
Google reported campaigns that used compromised credentials, fraudulent accounts and proxies to collect model outputs for distillation: training another model from a provider’s responses. A February 2026 reasoning-focused campaign exceeded 100,000 prompts; later campaigns included some above 100 million. Those are campaign query volumes, not model weights stolen or a measured replication success rate. Google says it disrupted accounts and reduced the usefulness of collected outputs.
Why it relates: Google’s compromised-account and fraudulent-account campaigns distributed repeated inference across account routes to acquire outputs for model cloning.
Google’s compromised-account and fraudulent-account campaigns; capability distillation is distinguished from stealing original weights. The prompt totals are collection volumes, not measured replication success or model weights. No autonomous agent deployment or protected-reasoning recovery is established.
Anthropic attributed more than 16 million exchanges through roughly 24,000 fraudulent accounts to DeepSeek, Moonshot AI and MiniMax. The flow distinguishes proxy access, authenticated accounts, distributed routing, inference-response collection and downstream model training. Provider-side hunting and verified limits act before outputs leave; potential harms from models lacking safeguards remain Anthropic’s risk assessment, not observed downstream victim counts.
Why it relates: Anthropic’s DeepSeek, Moonshot AI and MiniMax campaigns distributed repeated inference across account routes to acquire outputs for model cloning.
Anthropic’s DeepSeek, Moonshot AI and MiniMax campaigns; capability distillation is distinguished from stealing original weights. The 24,000-account and 16-million-exchange figures do not quantify recovered weights or observed downstream harms. Potential harms remain Anthropic’s assessment; an agent deployment is not established.
Sysdig observed stolen cloud credentials used to probe hosted LLM access and logging, then support reverse-proxy access that could be resold. The entry exploit, possession of credentials, tool preparation and provider-side use are distinct steps. The flow preserves parallel preparation before service use. Potential daily costs are estimates; provider-native IAM and complete invocation telemetry constrain direct calls that bypass an application proxy.
Why it relates: The Sysdig Threat Research Team discovered that malicious actors utilized stolen credentials to gain access to cloud-hosted large language models (LLMs). The relevant category is Unbounded Consumption: Unbounded consumption occurs when an LLM application does not adequately limit resource use, enabling excessive inference, unsustainable charges, service disruption, or model cloning.
The Sysdig Threat Research Team discovered that malicious actors utilized stolen credentials to gain access to cloud-hosted large language models (LLMs). The actors covertly gathered information about which models were enabled on the cloud service and created a reverse proxy for LLMs that would allow them to provide model access to cybercriminals. The Sysdig researchers identified tools used by the unknown actors that could target a broad range of cloud services including AI21 Labs, Anthropic, AWS Bedrock, Azure, ElevenLabs, MakerSuite, Mistral, OpenAI, OpenRouter, and GCP Vertex AI. Their technical analysis represented in the procedure below looked at at Amazon CloudTrail logs from the Amazon Bedrock service. The Sysdig researchers estimated that the worst-case financial harm for the unauthorized use of a single Claude 2.x model could be up to $46,000 a day. Update as of April 2025: This attack is ongoing and evolving. This case study only covers the initial reporting from Sysdig.
The publicly available Streamlit application MathGPT (https://mathgpt.streamlit.app/) uses GPT-3, a large language model (LLM), to answer user-generated math questions. Recent studies and experiments have shown that LLMs such as GPT-3 show poor performance when it comes to performing exact math directly. However, they can produce more accurate answers when asked to generate executable code that solves the question at hand. In the MathGPT application, GPT-3 is used to convert the user's natural language question into Python code that is then executed. After computation, the executed code and the answer are displayed to the user. Some LLMs can be vulnerable to prompt injection attacks, where malicious user inputs cause the models to perform unexpected behavior. In this incident, the actor explored several prompt-override avenues, producing code that eventually led to the actor gaining access to the application host system's environment variables and the application's GPT-3 API key, as well as executing a denial of service attack. As a result, the actor could have exhausted the application's API query budget or brought down the application. After disclosing the attack vectors and their results to the MathGPT and Streamlit teams, the teams took steps to mitigate the vulnerabilities, filtering on select prompts and rotating the API key.
Why it relates: The publicly available Streamlit application MathGPT (https://mathgpt.streamlit.app/) uses GPT-3, a large language model (LLM), to answer user-generated math questions. The relevant category is Unbounded Consumption: Unbounded consumption occurs when an LLM application does not adequately limit resource use, enabling excessive inference, unsustainable charges, service disruption, or model cloning.
The publicly available Streamlit application MathGPT (https://mathgpt.streamlit.app/) uses GPT-3, a large language model (LLM), to answer user-generated math questions. Recent studies and experiments have shown that LLMs such as GPT-3 show poor performance when it comes to performing exact math directly. However, they can produce more accurate answers when asked to generate executable code that solves the question at hand. In the MathGPT application, GPT-3 is used to convert the user's natural language question into Python code that is then executed. After computation, the executed code and the answer are displayed to the user. Some LLMs can be vulnerable to prompt injection attacks, where malicious user inputs cause the models to perform unexpected behavior. In this incident, the actor explored several prompt-override avenues, producing code that eventually led to the actor gaining access to the application host system's environment variables and the application's GPT-3 API key, as well as executing a denial of service attack. As a result, the actor could have exhausted the application's API query budget or brought down the application. After disclosing the attack vectors and their results to the MathGPT and Streamlit teams, the teams took steps to mitigate the vulnerabilities, filtering on select prompts and rotating the API key.
Zenity's honeypots saw operators point autonomous agents and LLM clients at exposed Ollama (a local LLM serving runtime) and LiteLLM (an AI gateway for routing LLM requests) backends. The defender's model server did not have to be exploited with malware; being reachable, unauthenticated, or protected only by a placeholder key was enough to turn it into attacker-funded inference infrastructure.
Why it relates: Zenity Labs observed three distinct operators use exposed Ollama or LiteLLM endpoints between March and May 2026 without exploiting the server software. The relevant category is Unbounded Consumption: Unbounded consumption occurs when an LLM application does not adequately limit resource use, enabling excessive inference, unsustainable charges, service disruption, or model cloning. An open inference service supplies unauthorized compute; the operators' attempted external attacks and lack of confirmed outcomes remain separate.
Zenity Labs observed three distinct operators use exposed Ollama or LiteLLM endpoints between March and May 2026 without exploiting the server software. One sent a 140,000-character Strix prompt and actively retried against a live French auction site before Zenity blocked completion; one staged HexStrike AI with 150 tool definitions but assigned no live target; and one sent 18 Codex-shaped requests carrying an anti-safety web-auditor persona that Zenity assessed as manual web reverse-engineering work without a confirmed target. The flow preserves those different evidence bounds instead of forcing all three into one impact path.