Visibility into AI Agents

Прозрачность в отношении ИИ-агентов
Emma Bluemke, David Krueger, Markus Anderljung, Noam Kolt, Alan Chan, Nitarshan Rajkumar, Carson Ezell, M. R. Kaufmann, Kevin Wei, Lewis Hammond, Herbie Bradley, Lennart Heim, Max Kaufmann
2024-06-03

AI agent governanceAI agentsactivity loggingagent identifiersreal-time monitoring
Increased delegation of commercial, scientific, governmental, and personal activities to AI agents—systems capable of pursuing complex goals with limited supervision—may exacerbate existing societal risks and introduce new risks. Understanding and mitigating these risks involves critically evaluating existing governance structures, revising and adapting these structures where needed, and ensuring accountability of key stakeholders. Information about where, why, how, and by whom certain AI agents are used, which we refer to as visibility, is critical to these objectives. In this paper, we assess three categories of measures to increase visibility into AI agents: agent identifiers, real-time monitoring, and activity logging. For each, we outline potential implementations that vary in intrusiveness and informativeness. We analyze how the measures apply across a spectrum of centralized through decentralized deployment contexts, accounting for various actors in the supply chain including hardware and software service providers. Finally, we discuss the implications of our measures for privacy and concentration of power. Further work into understanding the measures and mitigating their negative impacts can help to build a foundation for the governance of AI agents.
1
Each measure can be implemented with differing levels of intrusiveness and informativeness, creating distinct governance and privacy trade-offs.
2
Improving visibility may create privacy risks and concentrate power, requiring further research to mitigate these negative impacts.
3
The applicability of visibility measures varies across centralized and decentralized deployments and involves multiple hardware and software supply-chain actors.
4
The paper assesses three visibility measures for AI agents: agent identifiers, real-time monitoring, and activity logging.
5
Visibility into where, why, how, and by whom AI agents are used is critical for evaluating governance, accountability, and societal risks.

AI agents (systems capable of pursuing complex goals with limited supervision)

Measures for increasing visibility into AI agents, including agent identifiers, real-time monitoring, and activity logging, and their implications for privacy and concentration of power

Publication Details
Publication Date
2024-06-03
Journal
Publisher
ISSN
Cited by
298
Access Type
Author Information
Authors
Emma Bluemke
David Krueger
Markus Anderljung
Noam Kolt
Alan Chan
Nitarshan Rajkumar
Carson Ezell
M. R. Kaufmann
Kevin Wei
Lewis Hammond
Herbie Bradley
Lennart Heim
Max Kaufmann
Explore further
Open the scid.ai AI chat with a ready-made request: it will find papers on a similar topic and help build a literature review.
Find similar papers in the chat →
Make a presentation
100%