Realizing Agentic AI Value: A Systems Approach to Autonomy and Risk

Alex Reppel; Stefanie Beninger; Henry Robben; Onur Eken

doi:10.1177/00081256261486300

Realizing Agentic AI Value: A Systems Approach to Autonomy and Risk

Alex Reppel
, Stefanie Beninger
, Henry Robben
, Onur Eken

School of Business and Management

Research output: Contribution to journal › Article › peer-review

Abstract

Agentic Artificial Intelligence (Agentic AI) promises to transform organizations, yet many executives remain uncertain about how to capture value while minimizing the risks of these systems. This article’s Agentic AI Approach Matrix helps executives consider the level of autonomy given to such systems of autonomous AI agents, as well as the organizational mindset. Executives should adopt a mindset of systems thinking when approaching Agentic AI. By embracing systems thinking to drive value, executives are wise to execute the 5Es model to establish, explore, evaluate, and enhance Agentic AI while being aware of the emergence of risks, including opacity, misalignment, feedback loops, sovereignty, and costs.

Original language	English
Journal	California Management Review
Early online date	18 Sept 2026
DOIs	https://doi.org/10.1177/00081256261486300
Publication status	E-pub ahead of print - 18 Sept 2026

Keywords

Artificial Intelligence
Risk Management
Digital Transformation
Capabilities

Access to Document

10.1177/00081256261486300Licence: CC BY-NC

Cite this

@article{1d4a7af406bf4083aa5e075eca4183aa,

title = "Realizing Agentic AI Value: A Systems Approach to Autonomy and Risk",

abstract = "Agentic Artificial Intelligence (Agentic AI) promises to transform organizations, yet many executives remain uncertain about how to capture value while minimizing the risks of these systems. This article{\textquoteright}s Agentic AI Approach Matrix helps executives consider the level of autonomy given to such systems of autonomous AI agents, as well as the organizational mindset. Executives should adopt a mindset of systems thinking when approaching Agentic AI. By embracing systems thinking to drive value, executives are wise to execute the 5Es model to establish, explore, evaluate, and enhance Agentic AI while being aware of the emergence of risks, including opacity, misalignment, feedback loops, sovereignty, and costs.",

keywords = "Artificial Intelligence, Risk Management, Digital Transformation, Capabilities",

author = "Alex Reppel and Stefanie Beninger and Henry Robben and Onur Eken",

year = "2026",

month = sep,

day = "18",

doi = "10.1177/00081256261486300",

language = "English",

journal = "California Management Review",

issn = "0008-1256",

publisher = "SAGE Publications Ltd",

}

TY - JOUR

T1 - Realizing Agentic AI Value

T2 - A Systems Approach to Autonomy and Risk

AU - Reppel, Alex

AU - Beninger, Stefanie

AU - Robben, Henry

AU - Eken, Onur

PY - 2026/9/18

Y1 - 2026/9/18

N2 - Agentic Artificial Intelligence (Agentic AI) promises to transform organizations, yet many executives remain uncertain about how to capture value while minimizing the risks of these systems. This article’s Agentic AI Approach Matrix helps executives consider the level of autonomy given to such systems of autonomous AI agents, as well as the organizational mindset. Executives should adopt a mindset of systems thinking when approaching Agentic AI. By embracing systems thinking to drive value, executives are wise to execute the 5Es model to establish, explore, evaluate, and enhance Agentic AI while being aware of the emergence of risks, including opacity, misalignment, feedback loops, sovereignty, and costs.

AB - Agentic Artificial Intelligence (Agentic AI) promises to transform organizations, yet many executives remain uncertain about how to capture value while minimizing the risks of these systems. This article’s Agentic AI Approach Matrix helps executives consider the level of autonomy given to such systems of autonomous AI agents, as well as the organizational mindset. Executives should adopt a mindset of systems thinking when approaching Agentic AI. By embracing systems thinking to drive value, executives are wise to execute the 5Es model to establish, explore, evaluate, and enhance Agentic AI while being aware of the emergence of risks, including opacity, misalignment, feedback loops, sovereignty, and costs.

KW - Artificial Intelligence

KW - Risk Management

KW - Digital Transformation

KW - Capabilities

UR - https://cmr.berkeley.edu/2026/09/69-2-realizing-agentic-ai-value-a-systems-approach-to-autonomy-and-risk/

U2 - 10.1177/00081256261486300

DO - 10.1177/00081256261486300

M3 - Article

SN - 0008-1256

JO - California Management Review

JF - California Management Review

ER -

Realizing Agentic AI Value: A Systems Approach to Autonomy and Risk

Abstract

Keywords

Access to Document

Other files and links

CoSTAR: National R&D Lab for Creative Industries

Cite this

Realizing Agentic AI Value: A Systems Approach to Autonomy and Risk

Abstract

Keywords

Access to Document

Other files and links

Projects

CoSTAR: National R&D Lab for Creative Industries

Cite this