I am a Ph.D. student at the University of Connecticut (expected 2027), advised by Prof. Shiri Dori-Hacohen in the Reducing Information Ecosystem Threats (RIET) Lab and working closely with MIT's Algorithmic Alignment Group. My research focuses on what it takes for AI systems to reason well about moral issues.
My earlier work on quantifying misalignment (AAAI-25) addressed preference divergence between agents. My recent position papers argue that alignment requires more than matching preferences—AI systems also need to follow appropriate reasoning norms. I'm now building knowledge graph infrastructure for philosophical argumentation to support this broader vision of moral reasoning for AI.
Beyond research, I work on AI safety evaluation and policy. I have contributed to frontier model evaluations through the OpenAI Red Teaming Network and Nemesys Insights, and to governance discussions through the Wilson Center's Pathways to AI Policy program, as a Google Policy Fellow at CDT, and at NIST workshops. I also founded BEACON and led the 2025 Machine Ethics and Reasoning Workshop.
Position: Evaluations of AI Moral Reasoning Still Miss Half of the Picture
Kierans, A., Dutt, R., Rittichier, K., Dori-Hacohen, S., Ghosh, A.
ACL 2026 Workshop on Evaluating Evaluations (EvalEval)
[Paper]
Intelligence Is Not the Bottleneck: Structural Barriers to Automating Alignment Research
Kierans, A., Casper, S., Ghosh, A.
Workshop on Technical AI Governance Research (TAIGR) @ ICML 2026
[Preprint]
Position: Aligning AI Requires Automating Reasoning Norms
Kierans, A., Ghosh, A., Dori-Hacohen, S.
Submitted to NeurIPS 2026 Position Track
[Preprint]
Why LLMs Should Be Reasonably Morally Inconsistent
Stenseke, J., Kierans, A., Pres, I., Hadfield-Menell, D.
Pluralistic Alignment @ ICML 2026 Workshop
[Preprint]
Catastrophic Liability: Managing Systemic Risks in Frontier AI Development
Kierans, A., Ritticher, K., Sonsayar, U., Ghosh, A.
TAIS 2025 and EAAMO 2025
[Preprint]
Quantifying Misalignment Between Agents: Towards a Sociotechnical Understanding of Alignment
Kierans, A., Ghosh, A., Hazan, H., Dori-Hacohen, S.
AAAI 2025, Special Track on AI Alignment
[Paper] [Preprint]
Benchmarked Ethics: A Roadmap to AI Alignment, Moral Knowledge, and Control
Kierans, A.
AIES 2023
[Paper]
Bootstrap percolation via automated conjecturing
Bushaw, N., Conka, B., Gupta, V., Kierans, A., Lafayette, H., Larson, C., et al.
Ars Mathematica Contemporanea, 2023
[Paper]
See CV for complete history.
