R0044/2026-04-01/Q002/SRC05
Horowitz & Kahn (2024) — Bending the Automation Bias Curve in National Security
Source
| Field |
Value |
| Title |
Bending the Automation Bias Curve: A Study of Human and AI-Based Decision Making in National Security Contexts |
| Publisher |
International Studies Quarterly (Oxford Academic) |
| Author(s) |
Michael C. Horowitz (U. Penn), Lauren Kahn (Georgetown) |
| Date |
2024 |
| URL |
https://academic.oup.com/isq/article/68/2/sqae020/7638566 |
| Type |
Research paper (peer-reviewed) |
Summary
| Dimension |
Rating |
| Reliability |
High |
| Relevance |
Medium-High |
| Bias: Missing data |
Low risk |
| Bias: Measurement |
Low risk |
| Bias: Selective reporting |
Low risk |
| Bias: Randomization |
Low risk |
| Bias: Protocol deviation |
Low risk |
| Bias: COI/Funding |
Low risk |
Rationale
| Dimension |
Rationale |
| Reliability |
Published in a top IR journal. Large-scale experimental study with 9,000 respondents across 9 countries. |
| Relevance |
Directly measures automation bias in national security/military AI decision-making. Addresses military domain. However, focuses on automation bias (over-reliance on AI) rather than AI sycophancy (AI adjusting output to match user expectations). |
| Bias flags |
Well-controlled experimental design. Includes Dunning-Kruger effect analysis. |
| Evidence ID |
Summary |
| SRC05-E01 |
Quantified automation bias switching rates in military scenarios; Dunning-Kruger effect in AI reliance |