A recent Pacific Northwest National Laboratory webinar highlighted PermitAI, a federal artificial intelligence platform designed to help agencies speed up environmental reviews and permitting under the National Environmental Policy Act.
The platform includes tools to search historical NEPA documents, process public comments, draft sections of environmental review documents, and track public engagement opportunities. According to the webinar, PermitAI is being developed to support reviews for major infrastructure projects, including roads, bridges, mines, factories, power plants, and data centers.
Some of these tools could be useful, making thousands of historical NEPA documents easier to search could help agencies, advocates, researchers, and the public better understand past decisions. A public-facing tool that helps people find comment opportunities could also improve civic engagement.
At the same time, the webinar made clear that PermitAI is currently being tested primarily by federal agency users, while most public-facing tools are not yet available. The presenters also acknowledged that the system does not currently track whether NEPA documents have been litigated, found legally inadequate, or based on outdated information. It also does not determine whether a project complies with all relevant environmental laws or identify statutory gaps.
Historical NEPA documents are not neutral. They may contain outdated assumptions, weak mitigation language, incomplete cumulative impacts analysis, or inadequate treatment of subsistence, Tribal consultation, cultural resources, and environmental justice concerns. If AI tools rely heavily on past documents, they may reproduce old mistakes at a faster pace.
The public comment tool also deserves close scrutiny. PermitAI can sort, summarize, and group public comments, which may save agencies time. But public comments are not just data points to be processed. They are one of the few ways communities, Tribes, local residents, and advocates can speak directly into federal decisions that affect land, water, wildlife, and ways of life.
Federal permitting should not become faster at the expense of meaningful review.
As agencies begin using AI in environmental decision-making, they must commit to transparency, including disclosing when AI tools are used. They must ensure human experts remain responsible for legal and scientific judgments. They must protect Tribal consultation, Indigenous knowledge, cultural resources, and subsistence concerns from being oversimplified or automated. And they must make sure the public has access to the same information and opportunities as agencies and project proponents.
AI may change how environmental reviews are written and managed. It must not change the core purpose of NEPA: informed public decision-making before irreversible harm occurs.