Permitting Tech Plays: Operationalizing Rulings + Regulations

“We must recognize the limits of simply ordering change from the top rather than enabling change from all directions.”

— Jen Pahlka, Recoding America

A recent wave of historic federal investments in climate resilience, the clean energy transition, and new infrastructure means the government must deliver on a sprawling range of new projects tied to our national environmental goals. Those efforts will succeed or fail based on the government’s ability to quickly and effectively permit, site, build, and deploy clean energy infrastructure and restoration projects in the months and years ahead. The problem is that federal permitting—especially National Environmental Policy Act (NEPA) activities—constitutes a major barrier to the speed and scale of effort required to meet those goals. Environmental permitting is notoriously slow, opaque, unnecessarily resource-intensive, and often disconnected from clear, measurable outcomes rooted in NEPA’s original purpose; which was to meet the needs of present and future generations by minimizing damage to the environment, and by enriching our understanding of the vital natural systems and resources on which we all depend.

Finding and placing better, user-centered technology in the hands of key permitting staff can change that status quo.

That’s why, working with our partners at the Federation of American Scientists (FAS), we’re moving to take advantage of a once-in-a-generation opportunity for permitting innovation built around tangible user-needs—the thorny challenges plaguing agency staff, communities, and project proponents everyday across the permitting lifecycle—with an eye toward better outcomes over the long-term. And while permitting regulations, processes, and timelines span agency contexts and geographic regions, it’s clear that there are dedicated resources and growing momentum among policymakers, advocates, companies, and communities alike to fix permitting. What we need now are the levers of change and a roadmap for using them. This blog is the first in a series designed to provide both in the coming months.

From moving beyond obscure databases and static web maps, to demystifying the NEPA process and using tech built from the ground up—based on user needs—this series will sketch the permitting “tech plays” we think embody the principles at the heart of permitting innovation. Recognizing that technology isn’t a panacea for every barrier to better permitting, we’ll focus on action-oriented, concrete scenarios—user pain points and tech use cases that won’t shy away from in-the-weeds process and solutions. Our goal is to show how the work and experiences of NEPA practitioners, agency leadership, and project proponents can be improved. 

Play #1: Operationalizing rulings and regulations 

Those who make (and remake) federal policy, whether in Congress, via binding legal decisions, or deep within the ranks of agency offices, are rarely the same people who implement them. That is emphatically the case when it comes to  environmental permitting. Implementers are seldom lawyers or legal scholars—yet they are still charged in the course of their work with interpreting the many nuances and ambiguous directives built into NEPA (and related) permitting procedures. The consequences of that gap—between policy maker and implementer—leads to slow, and often inconsistent, adoption of new policies, including an increasingly complex, constantly shifting body of judicial rulings, agency policy, and administrative law. Fortunately, recent breakthroughs in artificial intelligence (AI) applications offer new opportunities for operationalizing permitting-related policy faster, more uniformly, and at lower cost to the taxpayer.

Status Quo: How policy and rulings are interpreted

To appreciate the potential of those tools to help NEPA practitioners operationalize rulings and regulations—to improve the process as well as permitting outcomes—consider today’s status quo when it comes to something like categorical exclusions (CatExs). By design, CatExs are meant to help accelerate the permitting by avoiding unnecessary NEPA procedures—that is, by excluding certain categories of projects from lengthy reviews if they’re deemed unnecessary given minimal anticipated impacts.

While that may sound like the purview of federal judges and lawyers dealing in obscure technicalities, CatExs are actually a strikingly clear example of where policy change, permitting processes, and shifts in legal interpretations intersect. [1] Whereas with other levels of NEPA review—Environmental Impact Statements (EISs) or Environmental Assessments (EAs) of proposed projects, for example—the merits of impact analysis, documentation, and project alternatives take center stage, they rarely affect interpretations of the review process itself. Judicial decisions about the criteria and applicability of specific CatExs, on the other hand, actually shape what qualifies as a project under specific categories—and those legal challenges and interpretations in turn shape how agencies attempt to scope and orchestrate projects. One common, and unfortunately, time-intensive, scenario typically unfolds like this:

• 1. A project is challenged in court.

  A project is contested by a litigant on the grounds that it doesn’t fit within the scope of a particular CatEx.

• 2. A judge makes a ruling.

  A judge makes a ruling. While the judge ruled in the project’s favor, the decision defines a more specific interpretation of the CatEx, stating in effect that, moving forward, every project using the category must address certain “new” elements to be compliant; the criteria change.

• 3. The agency’s national office then authors a briefing and distributes it.

  The agency’s national office—responsible for tracking the lawsuit—then authors a briefing and distributes it to all field staff offices, environmental coordinators, and agency NEPA planners.

• 4. The national office staff meet with their agency’s office of general counsel.

  In the ensuing months, the national office staff meet with their agency’s office of general counsel, seeking additional clarifications and working through potential scenarios where there may be legal gray areas. Once the national office staff have a stronger understanding of the ruling’s practical implications (how they will implement the ruling), they host a webinar for all environmental coordinators and NEPA planners to be briefed.

• 5. The planners are then made responsible for communicating their learnings to all local staff.

  The planners are then made responsible for communicating their learnings to all local staff affected by the changes, as well as all decision makers that must be accounted for in future CatExs.

• 6. Once units begin to review the revised CatEx, questions arise based on the new ruling.

  Once units begin to review the revised CatEx, questions arise based on the new ruling. Those questions are then sent back to the national office, and in some cases, onto the office of general counsel. Weeks—often months—later, answers to those inquiries arrive at the field office and adjustments to field staff’s procedures are made. 

• 7. Eight months later, the same project category is challenged again.

  Eight months later, the same project category is challenged again—or the agency revises its handbook to edit the relevant procedural language—and the whole process repeats itself.

The Problem + Cost: Low efficiency, low effectiveness

The basic problem with that scenario, beyond the fact of its ubiquity, is simple: it is a formula for low efficiency and high costs. For every change in NEPA implementing regulations issued by the Council on Environmental Quality (CEQ)—or in new case law, statute, Executive Orders, or agency procedures—NEPA planners must contend with a host of mutually reinforcing barriers as they attempt to implement. Although there are surely more, at minimum these include:

• 1. Slow information dissemination.

  Distributing important updates via email and subsequent webinars is slow and can often lead to delays in the implementation of new requirements. That lag also generates inconsistencies in how projects are managed across different, frequently siloed, field offices.

• 2. Communication breakdown.

  There are often several messengers or liaisons moving between new regulations or rulings and those tasked with implementing them. That game of “telephone” almost always creates numerous opportunities for misinterpretation or distortion of the original message.

• 3. Resource intensive processes.

  The process we’ve traced requires significant time and resources across multiple levels of an agency, including drafting briefings, organizing webinars, and extensive back-and-forth between teams for clarifications and guidance. The lack of a scaled, efficient approach translates into high personnel costs, and often, significantly delayed results.

• 4. Too much information.

  There is simply too much information for a single person—or even a single team—to account for in most permitting projects. As the scale of projects increase, so too do the chances of complex and lengthy litigation: across all federal agencies, projects developed under EISs are litigated 30% of the time; 40% of those cases ultimately lose, and that “losing” often takes 24 months or more. Depending on the details of the case, that proposed project might be revised, abandoned completely, or undergo additional analysis.

Solutions: Envisioning a Better Approach with AI

There are several areas where available technology can help upend that status quo by operationalizing new rulings and regulations—and in turn, improve efficiency across the permitting lifecycle. Consider these tech-enabled alternatives:

1. AI-Enabled Query Resolution. Develop a query resolution system where permitting staff can input questions and receive immediate answers based on a database of rulings, guidelines, and past projects. Recent developments in generative AI allow users to interact in a conversational way with an automated chat service trained on a particular dataset. A NEPA-focused (or broader environmental policy) query system could be built from resources like a centralized NEPA case repository, the Public Access to Court Electronic Records (PACER), environmental law organizations (Environmental Law Institute, Natural Resources Defense Council, Earthjustice); environmental law blogs and news (Environmental Law Reporter, SCOTUSblog); and academic institutions or research centers (Vermont Law School’s Environmental Law Center, or U.C., Berkeley’s Center for Law, Energy, and the Environment).

In June 2024, the Department of Energy’s (DOE) Pacific Northwest National Laboratory took the first steps in building out such a dataset with the release of NEPATEC1.0: the first large-scale text corpus of the National Environmental Policy Act. NEPATEC1.0 consists of textual data extracted from more than 28,000 EIS documents associated with 2,917 projects reviewed under NEPA. While this system would not be a panacea, it would address many of the high-cost, ineffective permitting practices we’ve highlighted here. For questions that an AI response cannot easily resolve, the application could simply escalate it to the appropriate human expert for review. Examples of policy-oriented applications that can be used as models “off-the-shelf” resources and solutions include BillTrack50, Thomson Reuters Practical Law AI, or NEPAccess. Broader generative AI applications include Claude, ChatGPT, and Llama.

Anthropic’s Claude allows users to upload documents, such as CEQ regulations, then generate queries specific to a user’s project or agency. 

2. AI Document Reviewer. Integrate an AI document reviewer into existing workflows that automatically checks all permits and analyses for compliance with the latest legal requirements. As documents are being created or updated, the AI can provide real-time feedback to ensure all necessary criteria (whatever their source) are addressed. In practice, those features can reduce the need for back-and-forth communication between teams and agency lawyers, as well as manual oversight. The AI can also detect errors, omissions, or inconsistencies in the documents and suggest corrections—enhancing accuracy and reducing the likelihood of non-compliance. (Including, for instance, through something like a litigation likelihood “score” based on past NEPA documents.) It could likewise learn from previous document reviews and continuously improve its accuracy and efficiency. Finally, the AI could also adapt to new legal interpretations and rulings as they are integrated into the system.

Juro AI Assistant responds to requests such as identifying legal risks and amending information based on legal guardrails.

3. Workflow and Permit Automation. Develop an activity-based system where legal updates linked to permitting procedures are integrated, and automatically applied, to all relevant analyses. This will ensure that every project uses the most current legal interpretations and requirements while minimizing traditional, often inconsistent, manual processing. Several automated workflows for local ordinances in the construction industry have been developed along these lines, but similar workflows for federal permits are less common. Relevant tools for improving permitting workflows include: PermitFlow, Camino, SchemeFlow AI, Symbium, and Spark.

Looking Ahead.

At a moment when the stakes for our national environmental and infrastructure efforts are increasingly existential, the status quo in permitting is untenable—and a major barrier to both the speed and scale of effort required to meet the moment. By closing gaps between regulators, judges, and policy implementers, AI can help NEPA practitioners across agencies operationalize new rulings and regulations, and in turn, recalibrate the permitting lifecycle to deliver more value—to our environment, communities, project leaders, and the agencies tasked with protecting our ecosystems and public health.

Stay tuned for subsequent blogs in this series, where we’ll explore other technology-enabled approaches to permitting innovation. Are you interested in EPIC’s restoration-focused permitting work? Read more about it here.

Something we missed? Questions or feedback for us about this work? We want to hear from you.

Endnotes:
1. See, for example, representative legal decisions here, here, and here. 
2. This series does not endorse specific software applications, including those employed in our examples–nor will we weigh in here on the ever-present question of “build versus buy” for government technology. Instead, we hope these alternatives inspire creative thinking around real-world examples of available tools, and a renewed dialogue on future possibilities for permitting innovation based on user needs.