Shawqi Al-Maliki

As large language models (LLMs) become deeply embedded in daily life, the urgent need for safer moderation systems that distinguish between naive and harmful requests while upholding appropriate censorship boundaries has never been greater. While existing LLMs can detect dangerous or unsafe content, they often struggle with nuanced cases such as implicit offensiveness, subtle gender and racial biases, and jailbreak prompts, due to the subjective and context-dependent nature of these issues. Furthermore, their heavy reliance on training data can reinforce societal biases, resulting in inconsistent and ethically problematic outputs. To address these challenges, we introduce GuardEval, a unified multi-perspective benchmark dataset designed for both training and evaluation, containing 106 fine-grained categories spanning human emotions, offensive and hateful language, gender and racial bias, and broader safety concerns. We also present GemmaGuard (GGuard), a Quantized Low-Rank Adaptation (QLoRA), fine-tuned version of Gemma3-12B trained on GuardEval, to assess content moderation with fine-grained labels. Our evaluation shows that GGuard achieves a macro F1 score of 0.832, substantially outperforming leading moderation models, including OpenAI Moderator (0.64) and Llama Guard (0.61). We show that multi-perspective, human-centered safety benchmarks are critical for mitigating inconsistent moderation decisions. GuardEval and GGuard together demonstrate that diverse, representative data materially improve safety, and adversarial robustness on complex, borderline cases.

Large Language Models (LLMs) have revolutionized the field of natural language processing. However, they exhibit some limitations, including a lack of reliability and transparency: they may hallucinate and fail to provide sources that support the generated output. Retrieval-Augmented Generation (RAG) was introduced to address such limitations in LLMs. One popular implementation, RAG-as-a-Service (RaaS), has shortcomings that hinder its adoption and accessibility. For instance, RaaS pricing is based on the number of submitted prompts, without considering whether the prompts are enriched by relevant chunks, i.e., text segments retrieved from a vector database, or the quality of the utilized chunks (i.e., their degree of relevance). This results in an opaque and less cost-effective payment model. We propose Chunk-as-a-Service (CaaS) as a transparent and cost-effective alternative. CaaS includes two variants: Open-Budget CaaS (OB-CaaS) and Limited-Budget CaaS (LB-CaaS), which is enabled by our “Utility-Cost Online Selection Algorithm (UCOSA)”. UCOSA further extends the cost-effectiveness and the accessibility of the OB-CaaS variant by enriching, in an online manner, a subset of the submitted prompts based on budget constraints and utility-cost tradeoff. Our experiments demonstrate the efficacy of the proposed UCOSA compared to both offline and relevance-greedy selection baselines. In terms of the performance metric—the number of enriched prompts (NEP) multiplied by the Average Relevance (AR)—UCOSA outperforms random selection by approximately 52% and achieves around 75% of the performance of offline selection methods. Additionally, in terms of budget utilization, LB-CaaS and OB-CaaS achieve higher performance-to-budget ratios of 140% and 86%, respectively, compared to RaaS, indicating their superior efficiency.

Deep Neural Networks (DNNs) have been the driving force behind many of the recent advances in machine learning. However, research has shown that DNNs are vulnerable to adversarial examples—input samples that have been perturbed to force DNN-based models to make errors. As a result, Adversarial Machine Learning (AdvML) has gained a lot of attention, and researchers have investigated these vulnerabilities in various settings and modalities. In addition, DNNs have also been found to incorporate embedded bias and often produce unexplainable predictions, which can result in anti-social AI applications. The emergence of new AI technologies that leverage Large Language Models (LLMs), such as ChatGPT and GPT-4, increases the risk of producing anti-social applications at scale. AdvML for Social Good (AdvML4G) is an emerging field that repurposes the AdvML bug to invent pro-social applications. Regulators, practitioners, and researchers should collaborate to encourage the development of pro-social applications and hinder the development of anti-social ones. In this work, we provide the first comprehensive review of the emerging field of AdvML4G. This paper encompasses a taxonomy that highlights the emergence of AdvML4G, a discussion of the differences and similarities between AdvML4G and AdvML, a taxonomy covering social good-related concepts and aspects, an exploration of the motivations behind the emergence of AdvML4G at the intersection of ML4G and AdvML, and an extensive summary of the works that utilize AdvML4G as an auxiliary tool for innovating pro-social applications. Finally, we elaborate upon various challenges and open research issues that require significant attention from the research community.

Deep neural networks have shown vulnerability to well-designed inputs called adversarial examples. Researchers in industry and academia have proposed many adversarial example defense techniques. However, they offer partial but not full robustness. Thus, complementing them with another layer of protection is a must, especially for mission-critical applications. This article proposes a novel online selection and relabeling algorithm (OSRA) that opportunistically utilizes a limited number of crowdsourced workers to maximize the machine learning (ML) system's robustness. The OSRA strives to use crowdsourced workers effectively by selecting the most suspicious inputs and moving them to the crowdsourced workers to be validated and corrected. As a result, the impact of adversarial examples gets reduced, and accordingly, the ML system becomes more robust. We also proposed a heuristic threshold selection method that contributes to enhancing the prediction system's reliability. We empirically validated our proposed algorithm and found that it can efficiently and optimally utilize the allocated budget for crowdsourcing. It is also effectively integrated with a state-of-the-art black box defense technique, resulting in a more robust system. Simulation results show that the OSRA can outperform a random selection algorithm by 60% and achieve comparable performance to an optimal offline selection benchmark. They also show that OSRA's performance has a positive correlation with system robustness.

Protecting the privacy of personal information, including emotions, is essential, and organizations must comply with relevant regulations to ensure privacy. Unfortunately, some organizations do not respect these regulations, or they lack transparency, leaving human privacy at risk. These privacy violations often occur when unauthorized organizations misuse machine learning (ML) technology, such as facial expression recognition (FER) systems. Therefore, researchers and practitioners must take action and use ML technology for social good to protect human privacy. One emerging research area that can help address privacy violations is the use of adversarial ML for social good. Evasion attacks, which are used to fool ML systems, can be repurposed to prevent misused ML technology, such as ML-based FER, from recognizing true emotions. By leveraging adversarial ML for social good, we can prevent organizations from violating human privacy by misusing ML technology, particularly FER systems, and protect individuals’ personal and emotional privacy. In this work, we propose an approach called Chaining of Adversarial ML Attacks (CAA) to create a robust attack that fools misused technology and prevents it from detecting true emotions. To validate our proposed approach, we conduct extensive experiments using various evaluation metrics and baselines. Our results show that CAA significantly contributes to emotional privacy preservation, with the fool rate of emotions increasing proportionally to the chaining length. In our experiments, the fool rate increases by 48% in each subsequent chaining stage of the chaining targeted attacks (CTA) while keeping the perturbations imperceptible (ε = 0.0001).

Deep Neural Networks (DDNs) have shown vulnerability to well-designed adversarial examples. Researchers in industry and academia have proposed many adversarial example defense techniques. However, none can provide complete robustness. The cutting-edge defense techniques offer partial reliability. Thus, complementing them with another layer of protection is a must, especially for mission-critical applications. This paper proposes a novel Online Selection and Relabeling Algorithm (OSRA) that opportunistically utilizes a limited number of crowdsourced workers to maximize the ML system’s robustness. OSRA strives to use crowdsourced workers effectively by selecting the most suspicious inputs and moving them to the crowdsourced workers to be validated and corrected. As a result, the impact of adversarial examples gets reduced, and accordingly, the ML system becomes more robust. We also proposed a heuristic threshold selection method that contributes to enhancing the prediction system’s reliability. We empirically validated our proposed algorithm and found that it can efficiently and optimally utilize the allocated budget for crowdsourcing. It is also effectively integrated with a state-of-the-art black-box defense technique, resulting in a more robust system. Simulation results show that OSRA can outperform a random selection algorithm by 60% and achieve comparable performance to an optimal offline selection benchmark. They also show that OSRA’s performance has a positive correlation with system robustness.