Slide
Centre Interdisciplinaire
de Recherche et d’Innovation
en Cybersécurité et Société
de Recherche et d’Innovation
en Cybersécurité et Société
1.
Elhajjout, A.; Houda, Z. A. E.; Moudoud, H.; Brik, B.; Jan, M. A.
Federated Large Language Models for A Trustworthy and Privacy-Preserving Healthcare: Applications, Challenges, and Future Research Directions Journal Article
In: IEEE Journal of Biomedical and Health Informatics, 2026, ISSN: 21682194 (ISSN).
Abstract | Links | BibTeX | Tags: Application research, Artificial intelligence, digital health, Distributed computer systems, Federated learning, Fine tuning, Health care, Health care application, Healthcare AI, human, Internet, Internet of medical thing, Internet of Medical Things, Language model, Large language model, large language models, Learning systems, male, Medical computing, natural language processing, Parameter-efficient fine-tuning, patient coding, Privacy, Privacy preservation, Privacy preserving, Privacy-preserving techniques, review, Sensitive data, trustworthiness
@article{elhajjoutFederatedLargeLanguage2026,
title = {Federated Large Language Models for A Trustworthy and Privacy-Preserving Healthcare: Applications, Challenges, and Future Research Directions},
author = {A. Elhajjout and Z. A. E. Houda and H. Moudoud and B. Brik and M. A. Jan},
url = {https://www.scopus.com/pages/publications/105034640126?origin=resultslist},
doi = {10.1109/JBHI.2026.3679612},
issn = {21682194 (ISSN)},
year = {2026},
date = {2026-01-01},
journal = {IEEE Journal of Biomedical and Health Informatics},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {Large-scale foundation models, especially Federated Large Language Models (FLLMs), aim to transform digital health by enabling clinically-grade natural language processing while keeping sensitive data local. However, their adoption is hindered by two main issues: (i) the computational and communication burden of parameter-rich models on resource-constrained Internet-of-Medical-Things (IoMT) devices, and (ii) performance degradation caused by Non-Independent and Identically Distributed (Non-IID) patient data. This paper presents a comprehensive survey of Federated Learning (FL) for LLMs in Healthcare (FedMed-LLMs). We review the foundations of FL and medical LLMs. Then, we present the FL-enabled LLMs applications in healthcare, and we examine their issues in terms of privacy, robustness, and trustworthiness. Finally, we present a set of core research problems and a comprehensive research agenda that identifies future directions for building robust and scalable FedMed-LLMs systems. © 2013 IEEE.},
keywords = {Application research, Artificial intelligence, digital health, Distributed computer systems, Federated learning, Fine tuning, Health care, Health care application, Healthcare AI, human, Internet, Internet of medical thing, Internet of Medical Things, Language model, Large language model, large language models, Learning systems, male, Medical computing, natural language processing, Parameter-efficient fine-tuning, patient coding, Privacy, Privacy preservation, Privacy preserving, Privacy-preserving techniques, review, Sensitive data, trustworthiness},
pubstate = {published},
tppubtype = {article}
}
Large-scale foundation models, especially Federated Large Language Models (FLLMs), aim to transform digital health by enabling clinically-grade natural language processing while keeping sensitive data local. However, their adoption is hindered by two main issues: (i) the computational and communication burden of parameter-rich models on resource-constrained Internet-of-Medical-Things (IoMT) devices, and (ii) performance degradation caused by Non-Independent and Identically Distributed (Non-IID) patient data. This paper presents a comprehensive survey of Federated Learning (FL) for LLMs in Healthcare (FedMed-LLMs). We review the foundations of FL and medical LLMs. Then, we present the FL-enabled LLMs applications in healthcare, and we examine their issues in terms of privacy, robustness, and trustworthiness. Finally, we present a set of core research problems and a comprehensive research agenda that identifies future directions for building robust and scalable FedMed-LLMs systems. © 2013 IEEE.
2.
Elhajjout, A.; Jarir, Z.; Moudoud, H.; Davoust, A.; Houda, Z. A. El
Leveraging Large Language Models for Contextual Threat Hypothesis Generation in IoT Networks Proceedings Article
In: Dig Tech Pap IEEE Int Conf Consum Electron, Institute of Electrical and Electronics Engineers Inc., 2026, ISBN: 0747668X (ISSN); 979-833155343-2 (ISBN), (Journal Abbreviation: Dig Tech Pap IEEE Int Conf Consum Electron).
Abstract | Links | BibTeX | Tags: Alert Triage, Cybersecurity, Hypotheses generation, Internet of thing security, Internet of things, IoT Security, Language model, Large datasets, Large language model, large language models, Network security, Prompt Engineering, Security alerts, Security operation center, Security Operations, Security systems, Threat Hypothesis Generation
@inproceedings{elhajjoutLeveragingLargeLanguage2026,
title = {Leveraging Large Language Models for Contextual Threat Hypothesis Generation in IoT Networks},
author = {A. Elhajjout and Z. Jarir and H. Moudoud and A. Davoust and Z. A. El Houda},
url = {https://www.scopus.com/pages/publications/105037351870?origin=resultslist},
doi = {10.1109/ICCE67443.2026.11449914},
isbn = {0747668X (ISSN); 979-833155343-2 (ISBN)},
year = {2026},
date = {2026-01-01},
booktitle = {Dig Tech Pap IEEE Int Conf Consum Electron},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {Modern Security Operations Centers face numerous challenges in managing the volume and complexity of security alerts from Internet of Things (IoT) networks. While traditional rule-based systems excel at detection, they provide limited contextual reasoning to help analysts understand ambiguous alerts. Large Language Models (LLMs) have shown promise across various cybersecurity domains. However, their potential to generate rich, explanatory threat hypotheses for ambiguous IoT alerts remains largely unexplored. To address these issues, in this paper, we present a systematic investigation of Large Language Model-based threat hypothesis generation with three key contributions. First, we introduce a context-aware structured prompting framework capable of synthesizing heterogeneous device telemetry into coherent threat narratives. Second, we formalize the assessment of explainability in security through a rigorous multidimensional quality metric system. Third, we provide the first comparative analysis of five state-of-the-art models on real-world IoT incidents, revealing critical performance tradeoffs. Results show that GPT-4o-mini achieves 88.2% accuracy on diverse attack types, while Qwen 3-235B achieves 95.2% accuracy on network-focused attacks. Statistical analysis reveals significant model-dataset interactions. These findings show that properly guided Large Language Models can augment analyst capabilities by providing detailed, contextual explanations that facilitate efficient alert triage. © 2026 IEEE.},
note = {Journal Abbreviation: Dig Tech Pap IEEE Int Conf Consum Electron},
keywords = {Alert Triage, Cybersecurity, Hypotheses generation, Internet of thing security, Internet of things, IoT Security, Language model, Large datasets, Large language model, large language models, Network security, Prompt Engineering, Security alerts, Security operation center, Security Operations, Security systems, Threat Hypothesis Generation},
pubstate = {published},
tppubtype = {inproceedings}
}
Modern Security Operations Centers face numerous challenges in managing the volume and complexity of security alerts from Internet of Things (IoT) networks. While traditional rule-based systems excel at detection, they provide limited contextual reasoning to help analysts understand ambiguous alerts. Large Language Models (LLMs) have shown promise across various cybersecurity domains. However, their potential to generate rich, explanatory threat hypotheses for ambiguous IoT alerts remains largely unexplored. To address these issues, in this paper, we present a systematic investigation of Large Language Model-based threat hypothesis generation with three key contributions. First, we introduce a context-aware structured prompting framework capable of synthesizing heterogeneous device telemetry into coherent threat narratives. Second, we formalize the assessment of explainability in security through a rigorous multidimensional quality metric system. Third, we provide the first comparative analysis of five state-of-the-art models on real-world IoT incidents, revealing critical performance tradeoffs. Results show that GPT-4o-mini achieves 88.2% accuracy on diverse attack types, while Qwen 3-235B achieves 95.2% accuracy on network-focused attacks. Statistical analysis reveals significant model-dataset interactions. These findings show that properly guided Large Language Models can augment analyst capabilities by providing detailed, contextual explanations that facilitate efficient alert triage. © 2026 IEEE.
3.
Sheehy, L.; Bouchard, S.; Kakkar, A.; Hakim, R. El; Lhoest, J.; Frank, A.
Development and Initial Testing of an Artificial Intelligence-Based Virtual Reality Companion for People Living with Dementia in Long-Term Care Journal Article
In: Journal of Clinical Medicine, vol. 13, no. 18, 2024, ISSN: 20770383 (ISSN), (Publisher: Multidisciplinary Digital Publishing Institute (MDPI)).
Abstract | Links | BibTeX | Tags: aged, article, Artificial intelligence, cognitive decline, cognitive impairment, compassion, conversation, dementia, Elderly, female, human, large language models, long term care, long-term care, major clinical study, male, program acceptability, program feasibility, reaction time, reminiscence, speech discrimination, very elderly, virtual reality
@article{sheehy_development_2024,
title = {Development and Initial Testing of an Artificial Intelligence-Based Virtual Reality Companion for People Living with Dementia in Long-Term Care},
author = {L. Sheehy and S. Bouchard and A. Kakkar and R. El Hakim and J. Lhoest and A. Frank},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85205071099&doi=10.3390%2fjcm13185574&partnerID=40&md5=844732ff858a0d5feb0a95a54093ad4d},
doi = {10.3390/jcm13185574},
issn = {20770383 (ISSN)},
year = {2024},
date = {2024-01-01},
journal = {Journal of Clinical Medicine},
volume = {13},
number = {18},
publisher = {Multidisciplinary Digital Publishing Institute (MDPI)},
abstract = {Background/Objectives: Feelings of loneliness are common in people living with dementia (PLWD) in long-term care (LTC). The goals of this study were to describe the development of a novel virtual companion for PLWD living in LTC and assess its feasibility and acceptability. Methods: The computer-generated virtual companion, presented using a head-mounted virtual reality display, was developed in two stages. In Stage 1, the virtual companion asked questions designed to encourage conversation and reminiscence. In Stage 2, more powerful artificial intelligence tools allowed the virtual companion to engage users in nuanced discussions on any topic. PLWD in LTC tested the application at each stage to assess feasibility and acceptability. Results: Ten PLWD living in LTC participated in Stage 1 (4 men and 6 women; average 82 years old) and Stage 2 (2 men and 8 women; average 87 years old). Session lengths ranged from 0:00 to 5:30 min in Stage 1 and 0:00 to 53:50 min in Stage 2. Speech recognition issues and a limited repertoire of questions limited acceptance in Stage 1. Enhanced conversational ability in Stage 2 led to intimate and meaningful conversations with many participants. Many users found the head-mounted display heavy. There were no complaints of simulator sickness. The virtual companion was best suited to PLWD who could engage in reciprocal conversation. After Stage 2, response latency was identified as an opportunity for improvement in future versions. Conclusions: Virtual reality and artificial intelligence can be used to create a virtual companion that is acceptable and enjoyable to some PLWD living in LTC. Ongoing innovations in hardware and software will allow future iterations to provide more natural conversational interaction and an enhanced social experience. © 2024 by the authors.},
note = {Publisher: Multidisciplinary Digital Publishing Institute (MDPI)},
keywords = {aged, article, Artificial intelligence, cognitive decline, cognitive impairment, compassion, conversation, dementia, Elderly, female, human, large language models, long term care, long-term care, major clinical study, male, program acceptability, program feasibility, reaction time, reminiscence, speech discrimination, very elderly, virtual reality},
pubstate = {published},
tppubtype = {article}
}
Background/Objectives: Feelings of loneliness are common in people living with dementia (PLWD) in long-term care (LTC). The goals of this study were to describe the development of a novel virtual companion for PLWD living in LTC and assess its feasibility and acceptability. Methods: The computer-generated virtual companion, presented using a head-mounted virtual reality display, was developed in two stages. In Stage 1, the virtual companion asked questions designed to encourage conversation and reminiscence. In Stage 2, more powerful artificial intelligence tools allowed the virtual companion to engage users in nuanced discussions on any topic. PLWD in LTC tested the application at each stage to assess feasibility and acceptability. Results: Ten PLWD living in LTC participated in Stage 1 (4 men and 6 women; average 82 years old) and Stage 2 (2 men and 8 women; average 87 years old). Session lengths ranged from 0:00 to 5:30 min in Stage 1 and 0:00 to 53:50 min in Stage 2. Speech recognition issues and a limited repertoire of questions limited acceptance in Stage 1. Enhanced conversational ability in Stage 2 led to intimate and meaningful conversations with many participants. Many users found the head-mounted display heavy. There were no complaints of simulator sickness. The virtual companion was best suited to PLWD who could engage in reciprocal conversation. After Stage 2, response latency was identified as an opportunity for improvement in future versions. Conclusions: Virtual reality and artificial intelligence can be used to create a virtual companion that is acceptable and enjoyable to some PLWD living in LTC. Ongoing innovations in hardware and software will allow future iterations to provide more natural conversational interaction and an enhanced social experience. © 2024 by the authors.
4.
Sheehy, L.; Bouchard, S.; Kakkar, A.; Hakim, R. El; Lhoest, J.; Frank, A.
Development and Initial Testing of an Artificial Intelligence-Based Virtual Reality Companion for People Living with Dementia in Long-Term Care Journal Article
In: Journal of Clinical Medicine, vol. 13, no. 18, 2024, ISSN: 20770383 (ISSN).
Abstract | Links | BibTeX | Tags: aged, article, Artificial intelligence, cognitive decline, cognitive impairment, compassion, conversation, dementia, Elderly, female, human, large language models, long term care, long-term care, major clinical study, male, program acceptability, program feasibility, reaction time, reminiscence, speech discrimination, very elderly, virtual reality
@article{sheehyDevelopmentInitialTesting2024,
title = {Development and Initial Testing of an Artificial Intelligence-Based Virtual Reality Companion for People Living with Dementia in Long-Term Care},
author = {L. Sheehy and S. Bouchard and A. Kakkar and R. El Hakim and J. Lhoest and A. Frank},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85205071099&doi=10.3390%2fjcm13185574&partnerID=40&md5=844732ff858a0d5feb0a95a54093ad4d},
doi = {10.3390/jcm13185574},
issn = {20770383 (ISSN)},
year = {2024},
date = {2024-01-01},
journal = {Journal of Clinical Medicine},
volume = {13},
number = {18},
publisher = {Multidisciplinary Digital Publishing Institute (MDPI)},
abstract = {Background/Objectives: Feelings of loneliness are common in people living with dementia (PLWD) in long-term care (LTC). The goals of this study were to describe the development of a novel virtual companion for PLWD living in LTC and assess its feasibility and acceptability. Methods: The computer-generated virtual companion, presented using a head-mounted virtual reality display, was developed in two stages. In Stage 1, the virtual companion asked questions designed to encourage conversation and reminiscence. In Stage 2, more powerful artificial intelligence tools allowed the virtual companion to engage users in nuanced discussions on any topic. PLWD in LTC tested the application at each stage to assess feasibility and acceptability. Results: Ten PLWD living in LTC participated in Stage 1 (4 men and 6 women; average 82 years old) and Stage 2 (2 men and 8 women; average 87 years old). Session lengths ranged from 0:00 to 5:30 min in Stage 1 and 0:00 to 53:50 min in Stage 2. Speech recognition issues and a limited repertoire of questions limited acceptance in Stage 1. Enhanced conversational ability in Stage 2 led to intimate and meaningful conversations with many participants. Many users found the head-mounted display heavy. There were no complaints of simulator sickness. The virtual companion was best suited to PLWD who could engage in reciprocal conversation. After Stage 2, response latency was identified as an opportunity for improvement in future versions. Conclusions: Virtual reality and artificial intelligence can be used to create a virtual companion that is acceptable and enjoyable to some PLWD living in LTC. Ongoing innovations in hardware and software will allow future iterations to provide more natural conversational interaction and an enhanced social experience. © 2024 by the authors.},
keywords = {aged, article, Artificial intelligence, cognitive decline, cognitive impairment, compassion, conversation, dementia, Elderly, female, human, large language models, long term care, long-term care, major clinical study, male, program acceptability, program feasibility, reaction time, reminiscence, speech discrimination, very elderly, virtual reality},
pubstate = {published},
tppubtype = {article}
}
Background/Objectives: Feelings of loneliness are common in people living with dementia (PLWD) in long-term care (LTC). The goals of this study were to describe the development of a novel virtual companion for PLWD living in LTC and assess its feasibility and acceptability. Methods: The computer-generated virtual companion, presented using a head-mounted virtual reality display, was developed in two stages. In Stage 1, the virtual companion asked questions designed to encourage conversation and reminiscence. In Stage 2, more powerful artificial intelligence tools allowed the virtual companion to engage users in nuanced discussions on any topic. PLWD in LTC tested the application at each stage to assess feasibility and acceptability. Results: Ten PLWD living in LTC participated in Stage 1 (4 men and 6 women; average 82 years old) and Stage 2 (2 men and 8 women; average 87 years old). Session lengths ranged from 0:00 to 5:30 min in Stage 1 and 0:00 to 53:50 min in Stage 2. Speech recognition issues and a limited repertoire of questions limited acceptance in Stage 1. Enhanced conversational ability in Stage 2 led to intimate and meaningful conversations with many participants. Many users found the head-mounted display heavy. There were no complaints of simulator sickness. The virtual companion was best suited to PLWD who could engage in reciprocal conversation. After Stage 2, response latency was identified as an opportunity for improvement in future versions. Conclusions: Virtual reality and artificial intelligence can be used to create a virtual companion that is acceptable and enjoyable to some PLWD living in LTC. Ongoing innovations in hardware and software will allow future iterations to provide more natural conversational interaction and an enhanced social experience. © 2024 by the authors.