SyncAI.news, a Varaisys broadcasting
ChronoPatch: Black-Box Energy-Latency Attacks on Vision-Language Models
JX

Jiawei Xing, Li Wang, Xiangtao Meng, Chuanchao Zang, Rui Wen, Zheng Li

· 1 min read

ResearcharXiv cs.AI

ChronoPatch: Black-Box Energy-Latency Attacks on Vision-Language Models

arXiv:2611.26117v1 Announce Type: cross Abstract: Vision-language models (VLMs) are increasingly deployed as generative inference services, where autoregressive decoding makes response length a key determinant of latency and computational cost. This creates an availability risk: adversarial visual inputs can prolong generation and increase serving overhead. Existing energy-latency attacks typically rely on gradients, logits, or token-level feedback, leaving open whether such risks persist when an attacker observes only the final response. We present ChronoPatch, a black-box attack that induces prolonged VLM generation requiring only the final response. It discovers visual delay anchors through response-length-guided evolutionary search, then minimizes perturbation magnitude through subspace-calibrated decision-based optimization. Across four open-source VLMs, ChronoPatch increases inference latency by 47.86%--293.34% and GPU energy consumption by 41.44%--249.70% under a 5% spatial budget. It also induces more repetitive outputs, as reflected by increased n-gram repetition ratios. On three commercial VLM APIs, ChronoPatch increases response length by 37.6%--81.5% and client-observed latency by 1.82x--2.29x under query budgets of 500 per API. These results reveal a generation-based availability risk that persists under final-response-only access and motivate generation-aware safeguards for deployed generative AI services.

Original source

This story was published by arXiv cs.AI and written by Jiawei Xing, Li Wang, Xiangtao Meng, Chuanchao Zang, Rui Wen, Zheng Li. SyncAI.news shows a preview; the complete article is on the publisher's site.

Read the full story on arxiv.org

Similar News