Agentic AI: a comprehensive survey of architectures, applications, and future directions
Mohamad Abou Ali, Fadi Dornaika, Jinan Charafeddine · 2025 · 158 citationsRead the paper
Abstract Agentic AI represents a transformative shift in artificial intelligence, but its rapid advancement has led to a fragmented understanding, often conflating modern neural systems with outdated symbolic models—a practice known as conceptual retrofitting . This survey cuts through this confusion by introducing a novel dual-paradigm framework that categorizes agentic systems into two distinct lineages: the symbolic/classical (relying on algorithmic planning and persistent state) and the neural/generative (leveraging stochastic generation and prompt-driven orchestration). Through a systematic PRISMA-based review of 90 studies (2018–2025), we provide a comprehensive analysis structured around this framework across three dimensions: (1) the theoretical foundations and architectural principles defining each paradigm; (2) domain-specific implementations in healthcare, finance, and robotics, demonstrating how application constraints dictate paradigm selection; and (3) paradigm-specific ethical and governance challenges, revealing divergent risks and mitigation strategies. Our analysis reveals that the choice of paradigm is strategic: symbolic systems dominate safety-critical domains (e.g., healthcare), while neural systems prevail in adaptive, data-rich environments (e.g., finance). Furthermore, we identify critical research gaps, including a significant deficit in governance models for symbolic systems and a pressing need for hybrid neuro-symbolic architectures. The findings culminate in a strategic roadmap arguing that the future of Agentic AI lies not in the dominance of one paradigm, but in their intentional integration to create systems that are both adaptable and reliable . This work provides the essential conceptual toolkit to guide future research, development, and policy toward robust and trustworthy hybrid intelligent systems.
What the full text shows
Theory or simulation- Identification of a 'paradigm-market fit' where symbolic/hybrid architectures dominate safety-critical domains and neural systems dominate data-rich adaptive domains. (Across healthcare, robotics (symbolic) and finance, education (neural))
- Validation of a dual-paradigm taxonomy (Symbolic/Classical vs. Neural/Generative) for classifying agentic AI systems. — 90 studies (Literature from 2018–2025)
- Identification of a governance imbalance where neural paradigm ethics are heavily researched while symbolic system governance is critically underexplored.
- Neural agents exhibit 'amnesia' across sessions due to context window limitations and statelessness. (Neural/Generative paradigm)
- Symbolic systems are brittle and fail catastrophically when encountering novel scenarios not covered by predefined rules. (Symbolic/Classical paradigm)
- Setup
- Review of 90 studies across IEEE Xplore, ACM Digital Library, arXiv, SpringerLink, ScienceDirect, and Google Scholar.
- How
- A systematic PRISMA-based literature review of 90 studies (2018-2025) using a novel dual-paradigm taxonomy to categorize architectures and applications.
- Limitations
- Proprietary nature of state-of-the-art neural systems leads to incomplete architectural and performance data. · Methodological heterogeneity across studies prevents direct cross-study benchmarking. · Rapid evolution of the neural paradigm may mean some very recent developments are missing. · Simplification required when assigning hybrid architectures to a single paradigm.
Summarized by Gemma 4 from the open-access full text · 9 Oct 2026. Read it