Not yet human-reviewed

This source record was written automatically by the research pipeline and has not yet been reviewed by the maintainer or integrated into a topic page. Treat it as a provisional summary and verify claims against the original source.

AI Agents vs. Agentic AI: A Conceptual Taxonomy, Applications and Challenges

Citation. Sapkota, R., Roumeliotis, K. I., & Karkee, M. (2026). AI Agents vs. Agentic AI: A Conceptual Taxonomy, Applications and Challenges. Information Fusion, 126, 103599. https://doi.org/10.1016/j.inffus.2025.103599

Summary. This conceptual review distinguishes tool- and task-oriented AI agents from agentic AI systems characterized by coordinated multi-agent autonomy, dynamic task decomposition, and persistent memory. It maps applications and technical challenges across the two paradigms, including hallucination, brittleness, emergent behavior, and coordination failures. The paper offers design-oriented remedies such as retrieval augmentation, tool use, memory architectures, and causal modeling, rather than evidence about human or organizational outcomes.

Key claims

• The review characterizes AI agents as modular LLM- or LIM-enabled systems for task-specific automation, while agentic AI coordinates multiple agents with dynamic decomposition, persistent memory, and greater autonomy.
• It contrasts AI-agent applications such as customer support, scheduling, and summarization with agentic-AI applications such as research automation, robotic coordination, and medical decision support.
• It identifies hallucination, prompt brittleness, limited planning, and weak causal reasoning as AI-agent limitations, and inter-agent misalignment, error propagation, emergent unpredictability, explainability, and scalability as added agentic-AI challenges.

Evidence & limitations

• This is a broad conceptual and literature review, not a primary empirical study; its taxonomy and solution proposals do not establish causal effects in workplaces or organizations.
• The ingested Drive PDF is arXiv:2505.10468v4; Crossref identifies a later peer-reviewed Information Fusion version (volume 126, 2026) with the canonical DOI used above.
• Although the review discusses governance and applications, its contribution is predominantly technical systems architecture and implementation; its fit with this wiki’s behavioral and organizational lens requires human review before synthesis.

Feeds

• agentic-organization-design
• work-redesign
• Owner ruling 2026-10-12: keep — “a helpful agentic work taxonomy that could be a potentially useful parallel to human work.” Synthesize it as a taxonomy of agentic work forms read against human work design, not as systems architecture; do not re-propose rejection.