Agent skills bundle instructions, reference data, and executable helpers that let a general agent perform specialized tasks. Hosted providers can keep these files secret while selling access to task results, making the skill itself a valuable target. Existing disclosure defenses can block requests that ask for the skill or reproduce its text, but they cannot block customers from submitting the ordinary tasks the service is built to complete. We present Daydreaming, an execution-only attack that steals a multi-file skill through black-box task interactions. The victim is never asked to reveal the skill or grade a reconstruction. Instead, Daydreaming adaptively creates crafted tasks whose results distinguish possible hidden behaviors. It tests individual behaviors, uses attacker-controlled shadow agents to choose a design, and completes each file using stored victim results and local execution checks. We formalize three nested threat levels of access as Differential, Trace, and Output, and focus on Output, where the attacker sees only the final response and returned files. Across 7 skills and 4 victim models, Daydreaming recovers 86.8% of the original skill's capability at Output, outperforming SigLeak by almost 4x. It produces installable skills using a median of 32 victim calls per skill even with disclosure defenses enabled. These results show that hiding skill files and filtering direct disclosure do not, by themselves, prevent functional reconstruction through normal use.
Khawaja Abaid Ullah, Mohammad Javad Khojastehcs.LG cs.CR
Knowledge distillation enables an adversary to replicate a proprietary classifier by querying its prediction interface and training a surrogate on the returned probability vectors. Antidistillation sampling, proposed for large language models, counters this threat with an input-dependent, gradient-directed perturbation of the served distribution; its transfer to classification has not been studied. Adapting the defense to classification, we show its behavior is governed by the distribution of the teacher's per-input confidence margins. Because well-trained classifiers are severely overconfident, the direct transfer exhibits an inert window: below a closed-form-predictable threshold, it affects neither attacker nor defender; beyond it, the defense undergoes a phase transition and degrades the teacher faster than the attacker's student. Temperature softening rescales the transition in closed form, and every temperature configuration lies on the same unfavorable trade-off curve. Our method, ADS-C, composes the perturbation under a closed-form, per-input margin budget that provably preserves every served top-1 prediction, so the defended teacher's accuracy equals the undefended teacher's identically. Under this guarantee the distilled student still loses 17.4 percentage points on CIFAR-100, 29.6 on CIFAR-10, and 13.3 on Tiny-ImageNet; matching this degradation with the unmodified defense costs 27.5, 32.9, and 22.2 points of teacher accuracy. Because served labels are unchanged, a hard-label attacker gains nothing, while the defended soft output trains a student up to 29.7 points below that floor: the incentive to distill served probabilities is not merely removed but reversed. To our knowledge, ADS-C is the first antidistillation defense for classification whose utility cost is exactly zero.
We present a black-box model-stealing attack that recovers private vision-tokenizer configurations of deployed vision-language models (VLMs), including the visual patch size and input preprocessing pipeline. The key idea is a task-level side channel induced by ViT-style patchification: when a synthetic grid image is aligned with the hidden patch grid, boundary cues are erased at tokenization, causing periodic accuracy drop. By sweeping the grid cell size and measuring these collapses, we infer the patch size; by introducing padding and a consistency-check test, we further identify whether preprocessing is dynamic- or fixed-resolution and recover the target resize resolution. Across open-source Qwen-VL variants and proprietary models including GPT and Claude, we reliably recover tokenizer-related parameters. Finally, we show that such leakage enables preprocessing-aware transfer attacks and model-targeted adversarial manipulation.
Model stealing attacks, where adversaries create high-fidelity surrogate models, are a significant threat to the intellectual property of machine learning services. Conventional wisdom suggests these surrogates could provide adversaries with economic leverage comparable to the original service providers. This paper challenges this assumption by evaluating model stealing attacks beyond mere fidelity to the target model. Because query-based extraction provides only partial supervision of the target's input-output behavior, the surrogate is not uniquely identified: many near-optimal surrogates can achieve comparable fidelity while differing in deployment-relevant properties. Instead of performing a classic learning-based model stealing attack, we compute the Rashomon Set (i.e., the set of almost-equally-accurate models) of surrogate models, and evaluate its diversity using multiplicity metrics (ambiguity, discrepancy, and Rashomon Capacity) and group fairness metrics. Across tabular, medical imaging, and NLP tasks, our experiments on real-world datasets reveal that despite exhibiting similar fidelity to the target model, surrogate models can display significant variances in other critical performance metrics. These findings cast doubt on the presumed equivalence between high-fidelity surrogates and the target model in practical deployment scenarios.
This paper uses geometry to explain how a machine learning model can be stolen using an already existing well-known method. The author has shown the exact conditions required to perfectly copy the final layer of a transformer network. When looking deeper into the hidden layers the author has explained clear limits. The author has also demonstrated that a hidden network cannot be fully reverse engineered just by looking at the final results. The research clearly maps out what can and cannot be stolen from a model.
Matthew Finlayson, Andreas Grivas, Xiang Ren +1cs.CR cs.AI cs.CC cs.CL
Language model parameters are known to impose unique (to each model) geometric constraints on their logit outputs, which serves as a signature that identifies the model, but also leaks the model's final layer parameters when an API distributes logits. We investigate more restrictive APIs that expose token rankings (i.e., their ordering by probability, but not the probability values) and find that rankings also constitute a signature: every model has a unique set of feasible top-$k$ rankings for sufficiently large $k$. Furthermore, the ranking signature is the first known (polynomially) unforgeable signature, since finding a model with the same set of feasible rankings is NP-hard. On the security front, we find that token rankings are already sufficient to approximately steal the final layer of the model, similar to logits, though the approximation is too coarse to forge the signature, and can be effectively countered by restricting the API to top-$k$ tokens with sufficiently small $k$. Since the top-$k$ required to present the model signature is generally smaller than the $k$ required to prevent stealing, it is possible for an API to present an unforgeable signature without leaking model parameters.
Maxime Schwarzer, Laurin Holz, Tobias Huerten +4cs.CR cs.AI
Artificial Intelligence (AI)-based Intrusion Detection Systems (IDS) deployed in energy infrastructure are vulnerable to model theft attacks, which allow adversaries to create evasive traffic offline. Current defences against model extraction rely either on identity-bound query monitoring, which is ineffective against distributed attackers (Sybil), or on prediction poisoning through soft-label perturbation, which is inapplicable to hard-label IDS deployments. Therefore, we propose FlowGuard, an identity-independent defence based on flow matching that classifies incoming queries as out-of-distribution (OOD) prior to IDS processing. This approach exploits the fact that queries generated synthetically for data-free model stealing attacks occupy a lower-dimensional manifold than real network traffic. This results in measurably lower log-likelihoods when using a Continuous Normalizing Flow that has been trained on legitimate data. We evaluate our method against PRADA and FDINet using MAZE and DisGUIDE attacks in single-client and distributed (100-client Sybil) settings. While PRADA's detection rate dropped to 0% when the distribution changed, our defence maintained a stable detection rate across both settings without relying on identity information. We discuss the scope and limitations of the approach, and outline potential applications to data-dependent attacks.
Closed-weight generative services are increasingly deployed through query-based APIs, where users can obtain generated outputs while model parameters remain inaccessible. However, such deployment does not prevent model stealing: an attacker can repeatedly query the service, collect large volumes of released synthetic images, and use them as training data for a private substitute model. This query-output-driven process enables unauthorized knowledge distillation and capability replication without direct access to the original weights. To mitigate this threat, a practical defense should preserve the visual fidelity of released images, provide explicit control over perturbation magnitude, and scale efficiently to large-volume output release. We present WaveGuard, a single-pass, generator-based protection framework that safeguards released synthetic images under a user-specified perturbation budget. WaveGuard employs a frequency-aware perturbation generator to inject structured, imperceptible perturbations that maintain perceptual utility for benign viewers while reducing the usefulness of protected images as training data for unauthorized student models. Extensive experiments under WikiArt-related synthetic-output distillation settings show that WaveGuard achieves a favorable efficacy--fidelity--efficiency trade-off, with explicit imperceptibility control and substantial gains in protection efficiency.