AI Initiatives
For Maine
About
AI Initiatives is a nonpartisan research group advancing responsible AI in Maine’s public sector. We are partnering with state agencies to run trials and develop governance tools that protect Mainers by pre-testing deployment outcomes.
We have reviewed the Maine AI Task Force’s preliminary report and identified several areas that require stronger language on safeguards. We have shared these findings with state officials and are pursuing strategic collaboration.
Climate
Policymakers studying generative AI (GenAI) are focusing on capability forecasts that overshadow evidence of user risk. For example, preliminary neurocognitive research has linked GenAI use to reduced functioning and education systems are adopting these tools without clearly defined safeguards. In industry, privacy standards and data practices for large technology rollouts are underspecified which puts professionals at heightened risk of information theft and data breach. These developments expose a risk gap in current GenAI policy which strengthens the case for research-led governance models.
Position
Maine GenAI adoption should be based on measurable performance outcomes. Upholding this position requires a governance model that will:
- Prioritize user-level cognitive outcomes and data privacy.
- Preregister rollout plans, risk thresholds, and pause/rollback triggers.
- Set ex ante deployment criteria with independent enforcement.
- Embed reversibility in procurement to manage evolving risk.
This approach requires testing user health, cognitive, and performance effects prior to public-sector rollout, treating GenAI as an intervention subject to clinical-style trials.
Research
Education
Characterize GenAI cognitive offloading among adolescents and examine patterns in academic performance and neurodevelopment.
Workplace
Study how GenAI adoption contributes to automation bias and skill degradation, assessing worker sentiment and conditions under varied use cases.
Tooling
Develop tools to measure how GenAI’s impacts on the state’s education and workforce systems evolve in response to distinct phases of adoption.
Impact
Our work alongside industry and state partners will result in AI research and legislation that puts the Maine public first. We have already drafted an NIH-style research agenda to serve as the foundation of this goal.
People
Nick Mathieu is a Topsham, Maine native trained in chemistry and protein science at Clark University. He is first author on papers in Frontiers in Oncology (2021) and Viruses (2021), contributor on NIH grant R15GM126432, and a student award recipient of the American Institute of Chemists. In industry, he has led business development research in regulated technology sectors and conducted due diligence under The Ablitt Group. Self-taught in pure and applied mathematics and an avid reader of philosophy, he now focuses on theory and methods for the ethical design and deployment of algorithmic systems.
Everett Gillis is a native of Brunswick, Maine trained in biochemistry at Gettysburg College. He worked as a teaching assistant under the Department of Chemistry at his alma mater while leading a 44-person STEM student organization and has advocated for humane technology deployment in public speaking. He has further experience as a researcher under the Department of Molecular Genetics at the University of Chicago. In 2023, he testified before the Joint Standing Committee on Health and Human Services to combat youth tobacco use. While he reads history and has contributed to additive combinatorics in his spare time, he is committed to improving health and learning outcomes for young students and Mainers.