The 2025 AI Agent Index: Documenting Technical and Safety Features of Deployed Agentic AI Systems

Leon Staufer; Kevin Feng; Kevin Wei; Luke Bailey; Yawen Duan; Mick Yang; A. Pinar Ozisik; Stephen Casper; Noam Kolt

doi:10.1145/3805689.3806728

The 2025 AI Agent Index: Documenting Technical and Safety Features of Deployed Agentic AI Systems

Leon Staufer^*
, Kevin Feng
, Kevin Wei
, Luke Bailey
, Yawen Duan
, Mick Yang
, A. Pinar Ozisik
, Stephen Casper
, Noam Kolt

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Agentic AI systems are increasingly capable of performing professional and personal tasks with limited human involvement. However, tracking these developments is difficult because the AI agent ecosystem is complex, rapidly evolving, and inconsistently documented, posing obstacles to both researchers and policymakers. To address these challenges, this paper presents the 2025 AI Agent Index. The Index documents information regarding the origins, design, capabilities, ecosystem, and safety features of 30 state-of-the-art AI agents based on publicly available information and email correspondence with developers. In addition to documenting information about individual agents, the Index illuminates broader trends in the development of agents, their capabilities, and the level of transparency of developers. Notably, we find that transparency varies substantially across agent developers and observe that most developers share little information about safety, evaluations, and societal impacts. The 2025 AI Agent Index is available online at https://aiagentindex.mit.edu.

Original language	English
Title of host publication	ACM FAccT 2026 - Proceedings of the 9th annual ACM Conference on Fairness, Accountability, and Transparency
Publisher	Association for Computing Machinery, Inc
Pages	1536-1576
Number of pages	41
ISBN (Electronic)	9798400725968
DOIs	https://doi.org/10.1145/3805689.3806728
State	Published - 25 Jun 2026
Event	9th Annual ACM Conference on Fairness, Accountability, and Transparency, ACM FAccT 2026 - Montreal, Canada Duration: 25 Jun 2026 → 28 Jun 2026

Publication series

Name	ACM FAccT 2026 - Proceedings of the 9th annual ACM Conference on Fairness, Accountability, and Transparency

Conference

Conference	9th Annual ACM Conference on Fairness, Accountability, and Transparency, ACM FAccT 2026
Country/Territory	Canada
City	Montreal
Period	25/06/26 → 28/06/26

Bibliographical note

Publisher Copyright:
© 2026 Copyright held by the owner/author(s).

Keywords

accountability
AI agent index
AI agents
ecosystem
sociotechnical systems
transparency

Access to Document

10.1145/3805689.3806728

Cite this

Staufer, L., Feng, K., Wei, K., Bailey, L., Duan, Y., Yang, M., Ozisik, A. P., Casper, S., & Kolt, N. (2026). The 2025 AI Agent Index: Documenting Technical and Safety Features of Deployed Agentic AI Systems. In ACM FAccT 2026 - Proceedings of the 9th annual ACM Conference on Fairness, Accountability, and Transparency (pp. 1536-1576). (ACM FAccT 2026 - Proceedings of the 9th annual ACM Conference on Fairness, Accountability, and Transparency). Association for Computing Machinery, Inc. https://doi.org/10.1145/3805689.3806728

Staufer, Leon ; Feng, Kevin ; Wei, Kevin et al. / The 2025 AI Agent Index : Documenting Technical and Safety Features of Deployed Agentic AI Systems. ACM FAccT 2026 - Proceedings of the 9th annual ACM Conference on Fairness, Accountability, and Transparency. Association for Computing Machinery, Inc, 2026. pp. 1536-1576 (ACM FAccT 2026 - Proceedings of the 9th annual ACM Conference on Fairness, Accountability, and Transparency).

@inproceedings{b5ae333d7d114bf7b372b779f7ad820d,

title = "The 2025 AI Agent Index: Documenting Technical and Safety Features of Deployed Agentic AI Systems",

abstract = "Agentic AI systems are increasingly capable of performing professional and personal tasks with limited human involvement. However, tracking these developments is difficult because the AI agent ecosystem is complex, rapidly evolving, and inconsistently documented, posing obstacles to both researchers and policymakers. To address these challenges, this paper presents the 2025 AI Agent Index. The Index documents information regarding the origins, design, capabilities, ecosystem, and safety features of 30 state-of-the-art AI agents based on publicly available information and email correspondence with developers. In addition to documenting information about individual agents, the Index illuminates broader trends in the development of agents, their capabilities, and the level of transparency of developers. Notably, we find that transparency varies substantially across agent developers and observe that most developers share little information about safety, evaluations, and societal impacts. The 2025 AI Agent Index is available online at https://aiagentindex.mit.edu.",

keywords = "accountability, AI agent index, AI agents, ecosystem, sociotechnical systems, transparency",

author = "Leon Staufer and Kevin Feng and Kevin Wei and Luke Bailey and Yawen Duan and Mick Yang and Ozisik, \{A. Pinar\} and Stephen Casper and Noam Kolt",

note = "Publisher Copyright: {\textcopyright} 2026 Copyright held by the owner/author(s).; 9th Annual ACM Conference on Fairness, Accountability, and Transparency, ACM FAccT 2026 ; Conference date: 25-06-2026 Through 28-06-2026",

year = "2026",

month = jun,

day = "25",

doi = "10.1145/3805689.3806728",

language = "אנגלית",

series = "ACM FAccT 2026 - Proceedings of the 9th annual ACM Conference on Fairness, Accountability, and Transparency",

publisher = "Association for Computing Machinery, Inc",

pages = "1536--1576",

booktitle = "ACM FAccT 2026 - Proceedings of the 9th annual ACM Conference on Fairness, Accountability, and Transparency",

}

Staufer, L, Feng, K, Wei, K, Bailey, L, Duan, Y, Yang, M, Ozisik, AP, Casper, S & Kolt, N 2026, The 2025 AI Agent Index: Documenting Technical and Safety Features of Deployed Agentic AI Systems. in ACM FAccT 2026 - Proceedings of the 9th annual ACM Conference on Fairness, Accountability, and Transparency. ACM FAccT 2026 - Proceedings of the 9th annual ACM Conference on Fairness, Accountability, and Transparency, Association for Computing Machinery, Inc, pp. 1536-1576, 9th Annual ACM Conference on Fairness, Accountability, and Transparency, ACM FAccT 2026, Montreal, Canada, 25/06/26. https://doi.org/10.1145/3805689.3806728

The 2025 AI Agent Index: Documenting Technical and Safety Features of Deployed Agentic AI Systems. / Staufer, Leon; Feng, Kevin; Wei, Kevin et al.
ACM FAccT 2026 - Proceedings of the 9th annual ACM Conference on Fairness, Accountability, and Transparency. Association for Computing Machinery, Inc, 2026. p. 1536-1576 (ACM FAccT 2026 - Proceedings of the 9th annual ACM Conference on Fairness, Accountability, and Transparency).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - The 2025 AI Agent Index

T2 - 9th Annual ACM Conference on Fairness, Accountability, and Transparency, ACM FAccT 2026

AU - Staufer, Leon

AU - Feng, Kevin

AU - Wei, Kevin

AU - Bailey, Luke

AU - Duan, Yawen

AU - Yang, Mick

AU - Ozisik, A. Pinar

AU - Casper, Stephen

AU - Kolt, Noam

PY - 2026/6/25

Y1 - 2026/6/25

N2 - Agentic AI systems are increasingly capable of performing professional and personal tasks with limited human involvement. However, tracking these developments is difficult because the AI agent ecosystem is complex, rapidly evolving, and inconsistently documented, posing obstacles to both researchers and policymakers. To address these challenges, this paper presents the 2025 AI Agent Index. The Index documents information regarding the origins, design, capabilities, ecosystem, and safety features of 30 state-of-the-art AI agents based on publicly available information and email correspondence with developers. In addition to documenting information about individual agents, the Index illuminates broader trends in the development of agents, their capabilities, and the level of transparency of developers. Notably, we find that transparency varies substantially across agent developers and observe that most developers share little information about safety, evaluations, and societal impacts. The 2025 AI Agent Index is available online at https://aiagentindex.mit.edu.

AB - Agentic AI systems are increasingly capable of performing professional and personal tasks with limited human involvement. However, tracking these developments is difficult because the AI agent ecosystem is complex, rapidly evolving, and inconsistently documented, posing obstacles to both researchers and policymakers. To address these challenges, this paper presents the 2025 AI Agent Index. The Index documents information regarding the origins, design, capabilities, ecosystem, and safety features of 30 state-of-the-art AI agents based on publicly available information and email correspondence with developers. In addition to documenting information about individual agents, the Index illuminates broader trends in the development of agents, their capabilities, and the level of transparency of developers. Notably, we find that transparency varies substantially across agent developers and observe that most developers share little information about safety, evaluations, and societal impacts. The 2025 AI Agent Index is available online at https://aiagentindex.mit.edu.

KW - accountability

KW - AI agent index

KW - AI agents

KW - ecosystem

KW - sociotechnical systems

KW - transparency

UR - https://www.scopus.com/pages/publications/105044408882

U2 - 10.1145/3805689.3806728

DO - 10.1145/3805689.3806728

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

AN - SCOPUS:105044408882

T3 - ACM FAccT 2026 - Proceedings of the 9th annual ACM Conference on Fairness, Accountability, and Transparency

SP - 1536

EP - 1576

BT - ACM FAccT 2026 - Proceedings of the 9th annual ACM Conference on Fairness, Accountability, and Transparency

PB - Association for Computing Machinery, Inc

Y2 - 25 June 2026 through 28 June 2026

ER -

Staufer L, Feng K, Wei K, Bailey L, Duan Y, Yang M et al. The 2025 AI Agent Index: Documenting Technical and Safety Features of Deployed Agentic AI Systems. In ACM FAccT 2026 - Proceedings of the 9th annual ACM Conference on Fairness, Accountability, and Transparency. Association for Computing Machinery, Inc. 2026. p. 1536-1576. (ACM FAccT 2026 - Proceedings of the 9th annual ACM Conference on Fairness, Accountability, and Transparency). doi: 10.1145/3805689.3806728

The 2025 AI Agent Index: Documenting Technical and Safety Features of Deployed Agentic AI Systems

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this