Tianqing Zhu

Youyang Qu, Ming Ding, Nan Sun et al.

Large Language Models (LLMs) are foundational to AI advancements, facilitating applications like predictive text generation. Nonetheless, they pose risks by potentially memorizing and disseminating sensitive, biased, or copyrighted information from their vast datasets. Machine unlearning emerges as a cutting-edge solution to mitigate these concerns, offering techniques for LLMs to selectively discard certain data. This paper reviews the latest in machine unlearning for LLMs, introducing methods for the targeted forgetting of information to address privacy, ethical, and legal challenges without necessitating full model retraining. It divides existing research into unlearning from unstructured/textual data and structured/classification data, showcasing the effectiveness of these approaches in removing specific data while maintaining model efficacy. Highlighting the practicality of machine unlearning, this analysis also points out the hurdles in preserving model integrity, avoiding excessive or insufficient data removal, and ensuring consistent outputs, underlining the role of machine unlearning in advancing responsible, ethical AI.

Dayong Ye, Tianqing Zhu, Congcong Zhu et al.

Machine unlearning refers to the process of mitigating the influence of specific training data on machine learning models based on removal requests from data owners. However, one important area that has been largely overlooked in the research of unlearning is reinforcement learning. Reinforcement learning focuses on training an agent to make optimal decisions within an environment to maximize its cumulative rewards. During the training, the agent tends to memorize the features of the environment, which raises a significant concern about privacy. As per data protection regulations, the owner of the environment holds the right to revoke access to the agent's training data, thus necessitating the development of a novel and pressing research field, known as \emph{reinforcement unlearning}. Reinforcement unlearning focuses on revoking entire environments rather than individual data samples. This unique characteristic presents three distinct challenges: 1) how to propose unlearning schemes for environments; 2) how to avoid degrading the agent's performance in remaining environments; and 3) how to evaluate the effectiveness of unlearning. To tackle these challenges, we propose two reinforcement unlearning methods. The first method is based on decremental reinforcement learning, which aims to erase the agent's previously acquired knowledge gradually. The second method leverages environment poisoning attacks, which encourage the agent to learn new, albeit incorrect, knowledge to remove the unlearning environment. Particularly, to tackle the third challenge, we introduce the concept of ``environment inference attack'' to evaluate the unlearning outcomes.

Wenhan Chang, Tianqing Zhu, Yu Zhao et al.

In the era of rapid generative AI development, interactions with large language models (LLMs) pose increasing risks of misuse. Prior research has primarily focused on attacks using template-based prompts and optimization-oriented methods, while overlooking the fact that LLMs possess strong unconstrained deceptive capabilities to attack other LLMs. This paper introduces a novel jailbreaking method inspired by the Chain-of-Thought mechanism. The attacker employs mission transfer to conceal harmful user intent within dialogue and generates a progressive chain of lure questions without relying on predefined templates, enabling successful jailbreaks. To further improve the attack's strength, we incorporate a helper LLM model that performs randomized narrative optimization over multi-turn interactions, enhancing the attack performance while preserving alignment with the original intent. We also propose a toxicity-based framework using third-party LLMs to evaluate harmful content and its alignment with malicious intent. Extensive experiments demonstrate that our method consistently achieves high attack success rates and elevated toxicity scores across diverse types of LLMs under black-box API settings. These findings reveal the intrinsic potential of LLMs to perform unrestricted attacks in the absence of robust alignment constraints. Our approach offers data-driven insights to inform the design of future alignment mechanisms. Finally, we propose two concrete defense strategies to support the development of safer generative models.

Hui Sun, Tianqing Zhu, Zhiqiu Zhang et al.

Deep generative models have gained much attention given their ability to generate data for applications as varied as healthcare to financial technology to surveillance, and many more - the most popular models being generative adversarial networks and variational auto-encoders. Yet, as with all machine learning models, ever is the concern over security breaches and privacy leaks and deep generative models are no exception. These models have advanced so rapidly in recent years that work on their security is still in its infancy. In an attempt to audit the current and future threats against these models, and to provide a roadmap for defense preparations in the short term, we prepared this comprehensive and specialized survey on the security and privacy preservation of GANs and VAEs. Our focus is on the inner connection between attacks and model architectures and, more specifically, on five components of deep generative models: the training data, the latent code, the generators/decoders of GANs/ VAEs, the discriminators/encoders of GANs/ VAEs, and the generated data. For each model, component and attack, we review the current research progress and identify the key challenges. The paper concludes with a discussion of possible future attacks and research directions in the field.

Xiaohan Hao, Wei Ren, Ruoting Xiong et al.

Federate learning can conduct machine learning as well as protect the privacy of self-owned training data on corresponding ends, instead of having to upload to a central trusted data aggregation server. In mobile scenarios, a centralized trusted server may not be existing, and even though it exists, the delay will not be manageable, e.g., smart driving cars. Thus, mobile federate learning at the edge with privacy-awareness is attracted more and more attentions. It then imposes a problem - after data are trained on a mobile terminal to obtain a learned model, how to share the model parameters among others to create more accurate and robust accumulative final model. This kind of model sharing confronts several challenges, e.g., the sharing must be conducted without a third trusted party (autonomously), and the sharing must be fair as model training (by training data)is valuable. To tackle the above challenges, we propose a smart contract and IPFS (Inter-Planetary File System) based model sharing protocol and algorithms to address the challenges. The proposed protocol does not rely on a trusted third party, where individual-learned models are shared/stored in corresponding ends. Conducted through extensive experiments, three main steps of the proposed protocol are evaluated. The average executive time of the three steps are 0.059s, 0.060s and 0.032s, demonstrating its efficiency.