Applying actionable recourse and cross-examination to sequential decision-making

Investigate how actionable recourse frameworks and adversarial cross-examination approaches can be adapted and applied to sequential decision-making systems to support explainability, user interaction, and correction.

Background

The authors note a relative lack of research on explainability and human-in-the-loop interventions for sequential decision-making compared to predictive systems. They point to conceptual gaps regarding how established practices from fairness in predictive settings—like actionable recourse and adversarial scrutiny—might extend to sequential decision scenarios.

Addressing these questions would help bridge methodological divides and potentially yield tools that enhance transparency and auditability for systems that plan and act over time.

References

Not only is there still foundational algorithmic work to be done, but there are also open conceptual questions such as how ideas of actionable recourse or cross-examination might be applied to this setting.

— Fairness and Sequential Decision Making: Limits, Lessons, and Opportunities  (2301.05753 - Nashed et al., 2023) in Section 6.6

Second, the reported risk reductions are model-predicted changes, not causal or off-policy validation of real intervention effects---a limitation shared with the algorithmic recourse literature, where recourse without the true structural causal model remains open .

— CARRE: Counterfactual Action Retrieval and Reason Evaluation for Explainable Churn Prescription  (2609.09766 - Kim et al., 9 Sep 2026) in Section 6, Discussion and Limitations, second limitation