The 2025 AI Agent Index: Documenting Technical and Safety Features of Deployed Agentic AI Systems
arXiv:2602.17753v1 Announce Type: cross 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 different transparency levels among agent developers and observe that most developers share little information about saf
arXiv:2602.17753v1 Announce Type: cross 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 different transparency levels among 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
Executive Summary
The 2025 AI Agent Index is a comprehensive documentation of the technical and safety features of 30 state-of-the-art AI agents, addressing the challenges posed by the complex and rapidly evolving AI agent ecosystem. The Index, compiled from publicly available information and direct correspondence with developers, highlights varying levels of transparency among developers, particularly regarding safety, evaluations, and societal impacts. This resource aims to aid researchers and policymakers in tracking developments and understanding broader trends in AI agent capabilities and design.
Key Points
- ▸ The AI Agent Index documents 30 state-of-the-art AI agents.
- ▸ Transparency levels vary significantly among developers.
- ▸ Most developers provide limited information on safety and societal impacts.
Merits
Comprehensive Documentation
The Index provides a thorough and systematic documentation of AI agents, which is crucial for researchers and policymakers.
Identifies Transparency Trends
The Index highlights the varying levels of transparency among developers, which is essential for understanding the current state of AI development.
Accessible Resource
The Index is available online, making it a valuable and accessible tool for the broader community.
Demerits
Limited Scope
The Index only covers 30 AI agents, which may not be representative of the entire AI agent ecosystem.
Dependence on Public Information
The Index relies heavily on publicly available information and developer correspondence, which may not always be accurate or complete.
Potential Bias
The selection of agents and the interpretation of their features may be subject to bias, which could affect the overall findings.
Expert Commentary
The 2025 AI Agent Index represents a significant step forward in documenting the rapidly evolving landscape of AI agents. The comprehensive nature of the Index provides a valuable resource for researchers, policymakers, and developers alike. However, the limitations inherent in relying on publicly available information and the potential for bias in the selection and interpretation of data must be acknowledged. The Index's findings on transparency and safety features underscore the need for greater accountability in AI development. As AI agents become increasingly capable of performing complex tasks, the insights provided by the Index will be crucial in shaping both practical applications and regulatory frameworks. The Index serves as a call to action for developers to enhance transparency and for policymakers to establish robust guidelines to ensure the safe and ethical development of AI agents.
Recommendations
- ✓ Expand the scope of the Index to include a more representative sample of AI agents.
- ✓ Develop standardized guidelines for transparency and safety in AI development to ensure consistency and reliability in documentation.