defense arXiv Feb 26, 2026 · 5w ago
Yanpei Guo, Wenjie Qu, Linyu Wu et al. · National University of Singapore · Nanyang Technological University +1 more
Auditing framework using verifiable computation to detect LLM provider fraud — model substitution, quantization abuse, token overbilling — with under 1% overhead
Output Integrity Attack nlp
Commercial large language models are typically deployed as black-box API services, requiring users to trust providers to execute inference correctly and report token usage honestly. We present IMMACULATE, a practical auditing framework that detects economically motivated deviations-such as model substitution, quantization abuse, and token overbilling-without trusted hardware or access to model internals. IMMACULATE selectively audits a small fraction of requests using verifiable computation, achieving strong detection guarantees while amortizing cryptographic overhead. Experiments on dense and MoE models show that IMMACULATE reliably distinguishes benign and malicious executions with under 1% throughput overhead. Our code is published at https://github.com/guo-yanpei/Immaculate.
llm transformer National University of Singapore · Nanyang Technological University · University of California
ML Security Papers