# The Perils of Agency: How Developers Perceive, Prioritize, and Address Risks in Agentic AI Products

arXiv: https://arxiv.org/abs/2606.15485 · HTML: https://mlsanity.com/paper/2606.15485/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Software engineering coding agents, one of 16 directions in LLM agents and tool use.
- Builds on 13 of its 47 references that are in the map.
- First posted on arXiv 13 Jun 2026.

## Placed in

- LLM agents and tool use / [Software engineering coding agents](https://mlsanity.com/agents/software-engineering-coding-agents/)

## Builds on (most cited first, 13 of 13)

- [Datasheets for Datasets](https://mlsanity.com/paper/1803.09010/) (2018)
- [Model Cards for Model Reporting](https://mlsanity.com/paper/1810.03993/) (2018)
- [Identifying the Risks of LM Agents with an LM-Emulated Sandbox](https://mlsanity.com/paper/2309.15817/) (2023)
- [Trustworthy AI: From Principles to Practices](https://mlsanity.com/paper/2110.01167/) (2021)
- [Agent Instructs Large Language Models to be General Zero-Shot Reasoners](https://mlsanity.com/paper/2310.03710/) (2023)
- [Deepfakes, Phrenology, Surveillance, and More! A Taxonomy of AI Privacy Risks](https://mlsanity.com/paper/2310.07879/) (2023)
- [Characterizing AI Agents for Alignment and Governance](https://mlsanity.com/paper/2504.21848/) (2025)
- [AHA!: Facilitating AI Impact Assessment by Generating Examples of Harms](https://mlsanity.com/paper/2306.03280/) (2023)
- [AI Ethics Principles in Practice: Perspectives of Designers and Developers](https://mlsanity.com/paper/2112.07467/) (2021)
- [Ethics in the Age of AI: An Analysis of AI Practitioners' Awareness and Challenges](https://mlsanity.com/paper/2307.10057/) (2023)
- [A Risk Ontology for Evaluating AI-Powered Psychotherapy Virtual Agents](https://mlsanity.com/paper/2505.15108/) (2025)
- [Why Johnny Can't Use Agents: Industry Aspirations vs. User Realities with AI Agents](https://mlsanity.com/paper/2509.14528/) (2025)
- [Privy: Envisioning and Mitigating Privacy Risks for Consumer-facing AI Product Concepts](https://mlsanity.com/paper/2509.23525/) (2025)

## Abstract (arXiv)

Agentic AI systems act autonomously, use tools, adapt to context, and operate in complex real-world environments. However, these same characteristics can create or exacerbate product risks. We studied how industry developers (n=35) perceive, prioritize, and address the risks in their agentic AI products. We found that developers' perceptions of risk were closely tied to the qualities that made the product agentic, such as autonomy, tool use, and usage in a real-world context. Developers prioritized product and business risks before considering downstream societal risks like job displacement and end-user privacy. This prioritization also impacted developers' ability and motivation to mitigate agentic risks. Finally, developers lacked mature controls for containing agentic risks, often relying on constraining the same characteristics that make agents useful: e.g., autonomy and goal complexity. These findings reveal a capability vs. risk control tension in agentic AI development: developers need to address risks that emerge from agentic capabilities, yet they currently have limited support for doing so without constraining agentic functionality.