Conference Description
Key Takeaways
- Half-day academic workshop examining cyber threats that exploit human behaviour rather than technical vulnerabilities
- Topics include social engineering, phishing, deepfakes, misinformation and AI-driven attacks
- Designed for cybersecurity researchers, security engineers, AI specialists and policy advisors from academia, industry and government
- Held at Sapienza University of Rome as part of ESORICS 2026
- Sponsored by the AI, Robotics and Cybersecurity Center (ARC) at Örebro University, Sweden
Introduction
The Workshop on Human-Targeted Cyber Threats and Defenses (HumSec’26) brings together researchers and practitioners to examine cybersecurity challenges that centre on human behaviour rather than system-level exploits. Taking place at Sapienza University of Rome as part of the ESORICS 2026 conference, this half-day workshop addresses the growing sophistication of attacks that manipulate perception, trust and decision-making. As artificial intelligence enables threat actors to scale social engineering campaigns with unprecedented efficiency, the need for cross-disciplinary collaboration between security specialists and human-factors researchers has become increasingly urgent.
About HumSec’26
HumSec’26 is an academic workshop with a technical and interdisciplinary orientation. The programme features keynote presentations, peer-reviewed paper sessions, poster displays and interactive discussions. By co-locating with ESORICS, one of Europe’s established security research conferences, the workshop provides participants with access to a broader community of cybersecurity scholars and practitioners.
The event is sponsored by the AI, Robotics and Cybersecurity Center (ARC) at Örebro University in Sweden, reflecting the workshop’s emphasis on the intersection of artificial intelligence and human-centric security research.
Human-Targeted Threats in an Era of AI-Enabled Attacks
Traditional cybersecurity has concentrated heavily on technical vulnerabilities—software flaws, misconfigurations and network weaknesses. Human-targeted threats operate differently, exploiting cognitive biases, emotional responses and the inherent trust that underpins organisational communication. Phishing remains the most prevalent example, but the category extends to sophisticated scams, online manipulation campaigns and coordinated disinformation efforts.
The emergence of large language models and generative AI has fundamentally altered the threat landscape. Attackers can now produce convincing phishing messages at scale, generate synthetic media for impersonation and automate social engineering tactics that previously required significant manual effort. These capabilities lower the barrier to entry for malicious actors while increasing the difficulty of detection for defenders.
HumSec’26 examines both sides of this equation. Sessions address AI-driven attack methodologies alongside defensive technologies such as phishing detection systems, deepfake identification tools and coercion-resistant authentication mechanisms. The workshop also considers how machine learning can be applied to identify manipulation patterns and flag suspicious communications before they reach end users.
Bridging Technical Solutions and Human Behaviour
A persistent challenge in cybersecurity is the gap between technically sound defences and their practical effectiveness when deployed among real users. Security mechanisms that prove cumbersome or confusing often fail because people circumvent them or make errors under pressure. HumSec’26 dedicates attention to usability research, examining how security tools are perceived and whether they function reliably in operational environments.
The workshop also explores stress training for incident response, recognising that human performance degrades under the time pressure and uncertainty that characterise active security incidents. Building organisational resilience requires not only robust technical controls but also personnel who can respond effectively when those controls are tested.
Who Should Attend
HumSec’26 is designed for professionals engaged with the human dimensions of cybersecurity. The audience includes academic researchers in security and human-computer interaction, security engineers developing detection and prevention systems, AI and machine learning specialists working on threat identification, and policy advisors shaping organisational or governmental security strategies. Chief information security officers and security leaders responsible for workforce training and awareness programmes will also find relevant material in the workshop’s examination of real-world applicability.
The Case for Interdisciplinary Collaboration
Addressing human-targeted threats effectively requires expertise that spans multiple disciplines. Technical security knowledge must be combined with insights from psychology, behavioural economics and communication studies. HumSec’26 positions itself as a venue for fostering these connections, bringing together researchers who might otherwise operate in separate academic and professional communities. The workshop’s emphasis on both attack research and defensive innovation reflects a recognition that understanding adversary techniques is essential to building robust countermeasures.