Post-Doc Research Associate - AI Credible Climate Mitigation

Position Title

Post-Doc Research Associate - AI for Credible Climate Mitigation

This position may be eligible for a hybrid work arrangement that may include a partially remote work location, consistent with System Office policy. UNC Chapel Hill employees are generally required to reside within a reasonable commuting distance of their assigned duty station. The Data-Driven Enviro Lab (DDL) is an interdisciplinary and international research initiative based at UNC’s Institute for the Environment that is redefining how data is used to tackle the world’s most pressing environmental challenges.

Analyzing the global environment at multiple geographical and jurisdictional levels – from cities, countries, and regions to local communities and corporations – we turn complex, large-scale, and often messy datasets into actionable insights that shape policy, drive accountability, and create real‑world impact. We operate with a commitment to evidence‑based, equitable solutions, building connections across disciplines, sectors, and geographies while giving special attention to underrepresented communities and data‑scarce regions in the Global South.

The DDL was founded and is led by Associate Professor Angel Hsu. The DDL at UNC‑Chapel Hill seeks a Postdoctoral Research Associate to join CLAIM: the Center for Climate Leadership and AI‑driven Integrity in Mitigation. CLAIM advances the credible use of artificial intelligence, especially generative AI and large language models (LLMs), to accelerate climate mitigation while safeguarding integrity in climate commitments.

We are looking for a researcher whose agenda sits at the intersection of AI and climate action, with a particular focus on sub‑national governments and non‑state actors (e.g., cities, regions, companies, investors, civil society). The ideal candidate will research how to use LLMs as tools for mitigation strategy and accountability and scrutinize them as systems that can reproduce bias, misinformation, or green washing unless carefully benchmarked.

Working with an interdisciplinary team spanning computer science, climate policy, law, and social science, the Postdoctoral Research Associate will:

• Develop, adapt, and evaluate LLM/genAI systems for climate mitigation intelligence, especially for tracking and assessing commitments and actions by cities, regions, and non‑state actors.
• Design and validate benchmarks, metrics, and auditing pipelines that test the accuracy, credibility, fairness, and robustness of LLM outputs related to climate targets, disclosures, and policy claims.
• Rigorously test genAI/ML models against misinformation, disinformation, and green washing, including stress‑testing models under adversarial, low‑resource, and multilingual settings.
• Study how generative AI shapes real‑world climate behavior, including impacts on corporate and sub‑national mitigation planning, transparency, and accountability.
• Contribute to CLAIM’s broader mission: data collection, methodological innovation, open tools, papers, and policy‑facing outputs that improve climate commitment integrity.

The position is initially for 1 year with the possibility of extension pending funding and performance.

PhD in Computer Science, Computing, Statistics or Data Science, or related disciplines.

• PhD in Computer Science, Computing, Statistics or Data Science, or related disciplines.
• Demonstrated hands‑on experience training, fine‑tuning, or adapting LLMs / generative AI models (e.g., instruction tuning, domain adaptation, RAG pipelines), with evidence in publications, open‑source code, or deployed tools.
• Strong experience in LLM/genAI evaluation and benchmarking, including methods for factuality/hallucination testing, robustness, calibration, bias/fairness, and/or adversarial stress‑testing.
• Solid programming and ML/NLP engineering skills in Python and ideally modern deep‑learning stacks (e.g., PyTorch/JAX, Hugging Face/Transformers, vector databases, agentic/RAG tooling), with ability to build reproducible research pipelines.
• Working knowledge of climate change…