August 14, 2025Open Access

Secure Engineering of Autonomous AI Agents: A Threat-Driven Development Framework

View Full Paper

Ask AI

Bookmark

Authors

TATanvir AhmedAmerican International University-Bangladesh SHSamiul HasanUniversity of Central Florida ASAhammed ShorifInternational Islamic University Chittagong

Discussion

Member takes

Implication

This exploration reveals security challenges of generative AI agents in organizations, highlighting critical vulnerabilities and mitigation strategies.

Key Points

The study identifies nine significant pitfalls in autonomous AI agents, indicating unique security challenges for organizations.
Notably, vulnerabilities include issues within the cognitive armature and trust boundaries, complicating standard security measures.
Assessment utilizes a comprehensive trouble model, proposing the advanced trouble framework and SHIELD for effective threat mitigation.
Comprehensive re-evaluation of threat models for generative AI is essential to prevent substantial liabilities in business settings.

Cite This Study

Ahmed et al. (2025) studied this question.

synapsesocial.com/papers/68af453fad7bf08b1ead2b99 https://doi.org/10.20944/preprints202508.1003.v1

View Full Paper

Ask AI

Bookmark

Implication

This exploration reveals security challenges of generative AI agents in organizations, highlighting critical vulnerabilities and mitigation strategies.

Secure Engineering of Autonomous AI Agents: A Threat-Driven Development Framework

Discussion

Member takes

Implication

Key Points

Cite This Study

Secure Engineering of Autonomous AI Agents: A Threat-Driven Development Framework

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider