The Legal AI Energy and Infrastructure Sector Adoption Report 2026: How Project Finance, Regulatory Permitting, and Grid Development Legal Teams Are Using — and Avoiding — AI Tools in the Most Document-Intensive Practice Group Nobody Talks About

Executive Summary

Energy and infrastructure legal teams sit at the intersection of the most document-intensive, jurisdiction-fragmented, and technically specialized practice areas in corporate law. Project finance counsel managing utility-scale solar transactions routinely process interconnection agreements, power purchase agreements, environmental impact statements, and state utility commission orders simultaneously — document sets that can exceed 50,000 pages per project. Regulatory counsel practicing before FERC and state PUCs navigate a patchwork of dockets, orders, and compliance obligations that change faster than most horizontal AI tools can track. Despite this, energy legal AI adoption has received almost none of the attention lavished on BigLaw M&A or litigation e-discovery use cases. That gap is closing, rapidly and unevenly. This briefing documents where adoption is real, where it is theater, and where the pressure — most acutely from FERC Order 2023's interconnection queue crisis — is forcing legal teams to move whether they are ready or not.

The FERC Order 2023 Pressure Point

No single regulatory development has done more to accelerate AI tool adoption in energy legal practice than FERC's Order 2023, issued in July 2023, which overhauled the interconnection process for new generation projects across the country. The order replaced the first-come, first-served serial queue with a "first-ready, first-served" cluster study process and imposed new readiness requirements that fundamentally changed how developers, utilities, and their counsel manage queue positions.

The practical consequence for legal teams has been brutal. As of early 2026, MISO's interconnection queue contains over 2,700 active projects representing more than 380 GW of proposed capacity. PJM's queue, despite its new cluster study process, still holds over 3,200 projects at various study stages. The legal work embedded in that backlog — interconnection agreement negotiations, study dispute filings, queue position transfer analyses, affected system analyses — is staggering. Outside counsel firms with significant independent power producer (IPP) client bases, including Norton Rose Fulbright, Troutman Pepper, and Stoel Rives, have all publicly acknowledged staffing and workflow challenges driven by queue volume.

The queue crisis has become the forcing function for AI adoption in energy legal that litigation volume was for e-discovery vendors a decade ago. Legal teams that previously ran interconnection queue monitoring on spreadsheets maintained by junior associates are actively piloting tools that can ingest ISO/RTO docket feeds, flag queue position changes, and generate summary alerts — because the alternative is missing a milestone that costs a client tens of millions in development capital.

What Tools Are Actually Deployed

The energy legal AI landscape in 2026 breaks into three tiers: horizontal AI platforms adapted for energy use, specialized energy legaltech tools built for this sector, and internal AI deployments at the largest utilities.

Horizontal platforms with meaningful energy adoption include Harvey, which has been adopted by multiple AmLaw 100 firms handling project finance transactions. At firms including Latham & Watkins and Vinson & Elkins, Harvey has been deployed for PPA redlining workflows. Internal reports from these deployments indicate meaningful time savings in first-pass redlining of standardized PPA provisions — particularly in negotiating warranty, curtailment, and change-in-law clauses where the platform can flag variance from firm-standard positions across hundreds of prior agreements. Latham's energy finance practice, which closes more than $20 billion annually in project finance transactions, began internal Harvey deployment in 2024 and has since expanded usage to include financing document comparison across complex tax equity and back-leverage structures.

Specialized energy tools are outperforming horizontal platforms in narrow but critical use cases. Gridline, a startup focused specifically on interconnection queue analytics, provides legal teams and project developers with real-time monitoring of ISO/RTO queue data, affected system study timelines, and queue position modeling. Law firms and IPP in-house teams using Gridline report that the tool has reduced the associate time spent on queue monitoring by approximately 60-70% per tracked project — a specific, auditable time saving that horizontal platforms cannot replicate because they lack the structured data feeds from MISO, PJM, CAISO, and SPP.

For environmental permitting, Lex Machina has been used by NEPA counsel to analyze Environmental Impact Statement challenges and litigation outcomes by circuit, project type, and agency. The tool does not automate NEPA compliance, but it substantially accelerates litigation risk assessment for projects facing third-party challenge. For permit condition tracking across multi-agency projects — where a single offshore wind development might carry conditions from BOEM, the Army Corps, EPA, USCG, and multiple state agencies simultaneously — firms including K&L Gates and Van Ness Feldman have begun using custom contract lifecycle management (CLM) tools configured to treat permit conditions as structured obligations with monitoring triggers, rather than narrative documents.

Utility in-house teams have moved more cautiously but more systematically. Eversource, Dominion Energy, and NextEra Energy's in-house legal operations teams have all disclosed AI initiatives in legal operations contexts. NextEra, which manages more wind and solar capacity than any other U.S. company and runs one of the largest in-house legal operations in the energy sector, has integrated AI document review into its PPA contract management workflow — though the company has been characteristically disciplined about disclosing operational details.

Where the Tools Are Failing

The failure modes are as instructive as the successes.

Jurisdictional regulatory variance remains the fundamental problem for horizontal AI platforms in energy regulatory practice. FERC practitioners can leverage AI for federal docket analysis with reasonable confidence. The moment that same analysis touches state PUC proceedings — which involve different procedural rules, different evidentiary standards, different docketing formats, and in many states, effectively no machine-readable data infrastructure — the tools break down. A Louisiana PSC general rate case, a California PUC integrated resource planning docket, and a New York PSC Article VII transmission siting proceeding have almost nothing structurally in common. No horizontal AI tool as of early 2026 has solved this fragmentation. Specialized practitioners at firms including Pillsbury Winthrop and Foley & Lardner consistently report that state commission work remains largely manual because the underlying data is not structured for machine consumption.

Interconnection queue priority disputes represent another failure zone. When a developer's queue position is challenged — through an affected system analysis dispute, a withdrawal cascade, or a cluster restudy that changes cost allocation — the legal analysis requires interpreting ISO/RTO tariff provisions, prior FERC orders, and project-specific interconnection study results simultaneously. This is precisely the kind of multi-source, technically specialized reasoning task where current AI tools hallucinate or produce confidently wrong outputs. Practitioners at Baker Botts and Stoel Rives have both flagged interconnection tariff interpretation as a use case where AI assistance has produced errors significant enough to require complete verification — essentially negating efficiency gains.

GC Pushes and Outside Counsel Guidelines

Energy sector GCs are beginning to push AI requirements into outside counsel guidelines, though adoption lags the broader corporate market. A 2025 survey by the Association of Corporate Counsel found that 34% of energy sector GCs had updated outside counsel guidelines to address AI use, compared to 51% in financial services. However, the energy GCs moving on this are moving substantively. NextEra's outside counsel guidelines, updated in 2025, include requirements that firms disclose which AI tools are used on NextEra matters and certify that AI-generated work product has been reviewed by a licensed attorney before delivery. Several large utilities have added AI audit rights provisions to their outside counsel engagement letters — a development that is pushing law firms to create more systematic AI use logging.

The inverse pressure is also appearing: several IPP clients of major energy law firms have begun asking whether their outside counsel are using AI tools, treating non-adoption as a potential efficiency and competitiveness concern. This is particularly true in project finance, where transaction timelines are compressed by interconnection deadlines and financing market windows.

Strategic Implications for Energy Legal Teams

The data from early 2026 points to a bifurcating market. Teams that have invested in specialized tools for specific high-volume, structured-data tasks — queue monitoring, PPA redlining, permit condition tracking — are generating measurable returns. Teams that have piloted horizontal AI platforms against unstructured regulatory documents without sector-specific configuration are generally reporting underwhelming results and reverting to manual workflows.

The most actionable posture for energy GCs and project finance partners is task-level specificity: identify the three or four document-intensive workflows that consume the most attorney hours per transaction or regulatory cycle, evaluate whether the underlying data is structured enough for current AI tools to process reliably, and pilot specialized tools against horizontal platforms in parallel before committing to enterprise deployment. The FERC Order 2023 backlog is not waiting for the technology to mature. The legal teams building systematic workflows now will have measurable capacity advantages over those still running queue monitoring on spreadsheets when the next wave of interconnection agreements hits closing tables.

The Legal Stack | Sector Adoption Research | 2026