mlsanityData to 7 Oct 2026

Membership Inference Attacks against Large Audio Language Models

arXiv 2603.28378 ↗First posted 30 Mar 2026cs.SD cs.AIcited by 1

On the map: LLM safety and security → Membership inference attacks

Abstract (arXiv)

We present the first systematic Membership Inference Attack (MIA) evaluation of LALMs. Using Multi-modal Blind Baselines based on textual, spectral and prosodic features, we demonstrate that common audio datasets exhibit near-perfect train/test separability (AUC ~ 1.0) even without model inference, thus MIA may primarily detect distribution shift. We therefore introduce a blind-baseline protocol to control for this confound. Under this protocol, we identify that the distribution-matched datasets enable reliable MIA evaluation without distribution-shift artifacts. We benchmark multiple MIA methods and conduct modality disentanglement experiments on these datasets. The results reveal that LALM memorization is cross-modal, arising only from binding a speaker's vocal identity with its text. These findings establish a principled standard for auditing LALMs beyond spurious correlations. Our codebase is available at https://github.com/snooow1029/ALM_MIA.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Placed in Membership inference attacks, one of 12 directions in LLM safety and security. Builds on 32 of its 37 references that are in the map. First posted on arXiv 30 Mar 2026.

Builds on

32 of 37 references.

Closest ancestors · 12

Same direction, newest first

  1. 2025Membership Inference Attacks Against Vision-Language Models27 Jan
  2. 2024EM-MIAs: Enhancing Membership Inference Attacks in Large Language Models through Ensemble Modeling23 Dec
  3. 2024Membership Inference Attacks against Large Vision-Language Models5 Nov
  4. 2024Scaling Up Membership Inference: When and How Attacks Succeed on Large Language Models31 Oct
  5. 2024Membership Inference Attacks Against In-Context Learning2 Sep
  6. 2024Noisy Neighbors: Efficient membership inference attacks against LLMs24 Jun
  7. 2024Blind Baselines Beat Membership Inference Attacks for Foundation Models23 Jun
  8. 2024LLM Dataset Inference: Did you train on my dataset?10 Jun
  9. 2024Do Membership Inference Attacks Work on Large Language Models?12 Feb
  10. 2023SLMIA-SR: Speaker-Level Membership Inference Attacks against Speaker Recognition Systems14 Sep
  11. 2023Membership Inference Attacks against Language Models via Neighbourhood Comparison29 May
  12. 2021Membership Inference Attacks Against Self-supervised Speech Models9 Nov

Other references · 20

  1. 2020Extracting Training Data from Large Language Modelscited by 2,144
  2. 2023Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Modelscited by 591
  3. 2023SALMONN: Towards Generic Hearing Abilities for Large Language Modelscited by 492
  4. 2021VoxPopuli: A Large-Scale Multilingual Speech Corpus for Representation Learning, Semi-Supervised Learning and Interpretationcited by 455
  5. 2021GigaSpeech: An Evolving, Multi-domain ASR Corpus with 10,000 Hours of Transcribed Audiocited by 391
  6. 2023Detecting Pretraining Data from Large Language Modelscited by 375
  7. 2023AudioPaLM: A Large Language Model That Can Speak and Listencited by 333
  8. 2024CosyVoice 2: Scalable Streaming Speech Synthesis with Large Language Modelscited by 282
  9. 2018TED-LIUM 3: twice as much data and corpus repartition for experiments on speaker adaptationcited by 224
  10. 2017Neural Audio Synthesis of Musical Notes with WaveNet Autoencoderscited by 222
  11. 2025Audio Flamingo 3: Advancing Audio Intelligence with Fully Open Large Audio Language Modelscited by 210
  12. 2025Audio Flamingo 2: An Audio-Language Model with Long-Audio Understanding and Expert Reasoning Abilitiescited by 127
8 more
  1. 2024Min-K%++: Improved Baseline for Detecting Pre-Training Data from Large Language Modelscited by 102
  2. 2021SPGISpeech: 5,000 hours of transcribed financial audio for fully formatted end-to-end speech recognitioncited by 77
  3. 2024TangoFlux: Super Fast and Faithful Text to Audio Generation with Flow Matching and Clap-Ranked Preference Optimizationcited by 67
  4. 2025DeSTA2.5-Audio: Toward General-Purpose Large Audio Language Model with Self-Generated Cross-Modal Alignmentcited by 47
  5. 2022CochlScene: Acquisition of acoustic scene data using crowdsourcingcited by 33
  6. 2025Music Flamingo: Scaling Music Understanding in Audio Language Modelscited by 19
  7. 2025Detecting Data Contamination in LLMs via In-Context Learningcited by 7
  8. 2025SeaLLMs-Audio: Large Audio-Language Models for Southeast Asiacited by 7

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Walking Through Uncertainty: An Empirical Study of Uncertainty Estimation for Audio-Aware Large Language Modelscited by 4