Yiming Li

System prompts are critical for shaping the behavior and output quality of large language model (LLM)-based applications, driving substantial investment in optimizing high-quality prompts beyond traditional handcrafted designs. However, as system prompts become valuable intellectual property, they are increasingly vulnerable to prompt theft and unauthorized use, highlighting the urgent need for effective copyright auditing, especially watermarking. Existing methods rely on verifying subtle logit distribution shifts triggered by a query. We observe that this logit-dependent verification framework is impractical in real-world content-only settings, primarily because (1) random sampling makes content-level generation unstable for verification, and (2) stronger instructions needed for content-level signals compromise prompt fidelity. To overcome these challenges, we propose PromptCOS, the first content-only system prompt copyright auditing method based on content-level output similarity. PromptCOS achieves watermark stability by designing a cyclic output signal as the conditional instruction's target. It preserves prompt fidelity by injecting a small set of auxiliary tokens to encode the watermark, leaving the main prompt untouched. Furthermore, to ensure robustness against malicious removal, we optimize cover tokens, i.e., critical tokens in the original prompt, to ensure that removing auxiliary tokens causes severe performance degradation. Experimental results show that PromptCOS achieves high effectiveness (99.3% average watermark similarity), strong distinctiveness (60.8% higher than the best baseline), high fidelity (accuracy degradation no greater than 0.6%), robustness (resilience against four potential attack categories), and high computational efficiency (up to 98.1% cost saving). Our code is available at GitHub (https://github.com/LianPing-cyber/PromptCOS).

The broad capabilities and substantial resources required to train Large Language Models (LLMs) make them valuable intellectual property, yet they remain vulnerable to copyright infringement, such as unauthorized use and model theft. LLM fingerprinting, a non-intrusive technique that compares the distinctive features (i.e., fingerprint) of LLMs to identify whether an LLM is derived from another, offers a promising solution to copyright auditing. However, its reliability remains uncertain due to the prevalence of diverse model modifications and the lack of standardized evaluation. In this SoK, we present the first comprehensive study of the emerging LLM fingerprinting. We introduce a unified framework and taxonomy that structures the field: white-box methods are classified based on their feature source as static, forward-pass, or backward-pass fingerprinting, while black-box methods are distinguished by their query strategy as either untargeted or targeted. Furthermore, we propose LeaFBench, the first systematic benchmark for evaluating LLM fingerprinting under realistic deployment scenarios. Built upon 7 mainstream foundation models and comprising 149 distinct model instances, LeaFBench integrates 13 representative post-development techniques, spanning both parameter-altering methods (e.g., fine-tuning, quantization) and parameter-independent techniques (e.g., system prompts, RAG). Extensive experiments on LeaFBench reveal the strengths and weaknesses of existing methods, thereby outlining future research directions and critical open problems in this emerging field. The code is available at https://github.com/shaoshuo-ss/LeaFBench.

The substantial investment required to develop Large Language Models (LLMs) makes them valuable intellectual property, raising significant concerns about copyright protection. LLM fingerprinting has emerged as a key technique to address this, which aims to verify a model's origin by extracting an intrinsic, unique signature (a "fingerprint") and comparing it to that of a source model to identify illicit copies. However, existing black-box fingerprinting methods often fail to generate distinctive LLM fingerprints. This ineffectiveness arises because black-box methods typically rely on model outputs, which lose critical information about the model's unique parameters due to the usage of non-linear functions. To address this, we first leverage Fisher Information Theory to formally demonstrate that the gradient of the model's input is a more informative feature for fingerprinting than the output. Based on this insight, we propose ZeroPrint, a novel method that approximates these information-rich gradients in a black-box setting using zeroth-order estimation. ZeroPrint overcomes the challenge of applying this to discrete text by simulating input perturbations via semantic-preserving word substitutions. This operation allows ZeroPrint to estimate the model's Jacobian matrix as a unique fingerprint. Experiments on the standard benchmark show ZeroPrint achieves a state-of-the-art effectiveness and robustness, significantly outperforming existing black-box methods.

LLM agents are highly vulnerable to Indirect Prompt Injection (IPI), where adversaries embed malicious directives in untrusted tool outputs to hijack execution. Most existing defenses treat IPI as an input-level semantic discrimination problem, which often fails to generalize to unseen payloads. We propose a new paradigm, action-level causal attribution, which secures agents by asking why a particular tool call is produced. The central goal is to distinguish tool calls supported by the user's intent from those causally driven by untrusted observations. We instantiate this paradigm with AttriGuard, a runtime defense based on parallel counterfactual tests. For each proposed tool call, AttriGuard verifies its necessity by re-executing the agent under a control-attenuated view of external observations. Technically, AttriGuard combines teacher-forced shadow replay to prevent attribution confounding, hierarchical control attenuation to suppress diverse control channels while preserving task-relevant information, and a fuzzy survival criterion that is robust to LLM stochasticity. Across four LLMs and two agent benchmarks, AttriGuard achieves 0% ASR under static attacks with negligible utility loss and moderate overhead. Importantly, it remains resilient under adaptive optimization-based attacks in settings where leading defenses degrade significantly.

Backdoor detection is currently the mainstream defense against backdoor attacks in federated learning (FL), where a small number of malicious clients can upload poisoned updates to compromise the federated global model. Existing backdoor detection techniques fall into two categories, passive and proactive, depending on whether the server proactively intervenes in the training process. However, both of them have inherent limitations in practice: passive detection methods are disrupted by common non-i.i.d. data distributions and random participation of FL clients, whereas current proactive detection methods are misled by an inevitable out-of-distribution (OOD) bias because they rely on backdoor coexistence effects. To address these issues, we introduce a novel proactive detection method dubbed Coward, inspired by our discovery of multi-backdoor collision effects, in which consecutively planted, distinct backdoors significantly suppress earlier ones. Correspondingly, we modify the federated global model by injecting a carefully designed backdoor-collided watermark, implemented via regulated dual-mapping learning on OOD data. This design not only enables an inverted detection paradigm compared to existing proactive methods, thereby naturally counteracting the adverse impact of OOD prediction bias, but also introduces a low-disruptive training intervention that inherently limits the strength of OOD bias, leading to significantly fewer misjudgments. Extensive experiments on benchmark datasets show that Coward achieves state-of-the-art detection performance, effectively alleviates OOD prediction bias, and remains robust against potential adaptive attacks. The code for our method is available at https://github.com/still2009/cowardFL.

Text-to-Image (T2I) models, represented by DALL$\cdot$E and Midjourney, have gained huge popularity for creating realistic images. The quality of these images relies on the carefully engineered prompts, which have become valuable intellectual property. While skilled prompters showcase their AI-generated art on markets to attract buyers, this business incidentally exposes them to \textit{prompt stealing attacks}. Existing state-of-the-art attack techniques reconstruct the prompts from a fixed set of modifiers (i.e., style descriptions) with model-specific training, which exhibit restricted adaptability and effectiveness to diverse showcases (i.e., target images) and diffusion models. To alleviate these limitations, we propose Prometheus, a training-free, proxy-in-the-loop, search-based prompt-stealing attack, which reverse-engineers the valuable prompts of the showcases by interacting with a local proxy model. It consists of three innovative designs. First, we introduce dynamic modifiers, as a supplement to static modifiers used in prior works. These dynamic modifiers provide more details specific to the showcases, and we exploit NLP analysis to generate them on the fly. Second, we design a contextual matching algorithm to sort both dynamic and static modifiers. This offline process helps reduce the search space of the subsequent step. Third, we interact with a local proxy model to invert the prompts with a greedy search algorithm. Based on the feedback guidance, we refine the prompt to achieve higher fidelity. The evaluation results show that Prometheus successfully extracts prompts from popular platforms like PromptBase and AIFrog against diverse victim models, including Midjourney, Leonardo.ai, and DALL$\cdot$E, with an ASR improvement of 25.0\%. We also validate that Prometheus is resistant to extensive potential defenses, further highlighting its severity in practice.

Vision Language Model (VLM)-based agents are stateful, autonomous entities capable of perceiving and interacting with their environments through vision and language. Multi-agent systems comprise specialized agents who collaborate to solve a (complex) task. A core security property is robustness, stating that the system should maintain its integrity under adversarial attacks. However, the design of existing multi-agent systems lacks the robustness consideration, as a successful exploit against one agent can spread and infect other agents to undermine the entire system's assurance. To address this, we propose a new defense approach, Cowpox, to provably enhance the robustness of multi-agent systems. It incorporates a distributed mechanism, which improves the recovery rate of agents by limiting the expected number of infections to other agents. The core idea is to generate and distribute a special cure sample that immunizes an agent against the attack before exposure and helps recover the already infected agents. We demonstrate the effectiveness of Cowpox empirically and provide theoretical robustness guarantees.