WHAT THIS BILL REGULATES · 3 REQUIREMENT TYPES
How Is This Bill Enforced
Verbatim statutory text on the left; plain-language analysis and a per-section checklist on the right. Numbered markers cross-link to the matching checklist row.
(1)–(4) The General Assembly makes the following findings: (1) Environmental permitting by the Department of Environmental Quality is a critical function of State government that protects public health, the environment, and the economy of North Carolina. (2) Advances in artificial intelligence, including large language model systems and retrieval-augmented generation tools, offer significant potential to assist agency staff in the drafting and review of environmental permitsEnvironmental permitEnvironmental permit. – Any permit, certificate, approval, or authorization issued by the Commission or the Department under this Article, Article 4 of Chapter 113A, Articles 9 or 10 of Chapter 130A, or Articles 21, 21A, or 21B of Chapter 143 of the General Statutes, or the rules adopted thereunder.G.S. § 143B-279.22(a)(4) by accelerating routine tasks, improving consistency, and reducing processing times. (3) The use of artificial intelligence as a permit drafting and permit review tool, subject to independent human review and decision-making by agency staff exercising professional judgement, is consistent with the principles of government efficiency. (4) A joint initiative between the Environmental Management CommissionCommissionCommission. – The Environmental Management Commission.G.S. § 143B-279.22(a)(2) and the Department of Environmental Quality to use artificial intelligence tools for environmental permitting will benefit permit applicants, regulated entities, State employees, and the general public.
Section 1 sets out the General Assembly's findings motivating the bill. It identifies environmental permitting as a critical state function and recognizes that AI tools — including large language models and retrieval-augmented generation — can improve consistency and reduce processing times when used subject to independent human review. No operative obligations are created by these findings.
(a)(1)–(4) Definitions. – The following definitions apply in this section: (1) Artificial intelligence or A.I.Artificial intelligence or A.I.Artificial intelligence or A.I. – A machine-based system that processes inputs and generates outputs, such as texts, recommendations, comparisons, summaries, and predictions. Artificial intelligence includes large language models, natural language processing tools, and retrieval-augmented generation systems.G.S. § 143B-279.22(a)(1) – A machine-based system that processes inputs and generates outputs, such as texts, recommendations, comparisons, summaries, and predictions. Artificial intelligence includes large language models, natural language processing tools, and retrieval-augmented generation systems. (2) CommissionCommissionCommission. – The Environmental Management Commission.G.S. § 143B-279.22(a)(2). – The Environmental Management CommissionCommissionCommission. – The Environmental Management Commission.G.S. § 143B-279.22(a)(2). (3) DepartmentDepartmentDepartment. – The Department of Environmental Quality.G.S. § 143B-279.22(a)(3). – The Department of Environmental Quality. (4) Environmental permitEnvironmental permitEnvironmental permit. – Any permit, certificate, approval, or authorization issued by the Commission or the Department under this Article, Article 4 of Chapter 113A, Articles 9 or 10 of Chapter 130A, or Articles 21, 21A, or 21B of Chapter 143 of the General Statutes, or the rules adopted thereunder.G.S. § 143B-279.22(a)(4). – Any permit, certificate, approval, or authorization issued by the CommissionCommissionCommission. – The Environmental Management Commission.G.S. § 143B-279.22(a)(2) or the DepartmentDepartmentDepartment. – The Department of Environmental Quality.G.S. § 143B-279.22(a)(3) under this Article, Article 4 of Chapter 113A, Articles 9 or 10 of Chapter 130A, or Articles 21, 21A, or 21B of Chapter 143 of the General Statutes, or the rules adopted thereunder.
Subsection (a) establishes the four defined terms used throughout the new statute section: artificial intelligence, Commission, Department, and environmental permit. The AI definition is intentionally broad, encompassing any machine-based system that processes inputs and generates outputs including texts, recommendations, comparisons, summaries, and predictions. No operative obligations are created by this subsection.
(b) 1 Joint A.I. Environmental Permitting Program; Established. – The CommissionCommissionCommission. – The Environmental Management Commission.G.S. § 143B-279.22(a)(2) and the DepartmentDepartmentDepartment. – The Department of Environmental Quality.G.S. § 143B-279.22(a)(3) shall jointly establish a program to implement the use of artificial intelligence to assist agency staff in the drafting and review of applications for environmental permitsEnvironmental permitEnvironmental permit. – Any permit, certificate, approval, or authorization issued by the Commission or the Department under this Article, Article 4 of Chapter 113A, Articles 9 or 10 of Chapter 130A, or Articles 21, 21A, or 21B of Chapter 143 of the General Statutes, or the rules adopted thereunder.G.S. § 143B-279.22(a)(4).
Subsection (b) is the bill's operative establishment clause. It directs the Environmental Management Commission and the Department of Environmental Quality to jointly create a program implementing AI to assist agency staff in drafting and reviewing environmental permit applications. The obligation runs exclusively to the two state entities — no private-sector compliance duty is created.
(c)(1)–(6) 2 Rulemaking. – The CommissionCommissionCommission. – The Environmental Management Commission.G.S. § 143B-279.22(a)(2) may adopt rules to establish standards and procedures for the use of artificial intelligence systems by agency staff of the Department of Environmental Quality to review and analyze permit applications and to draft environmental permitsEnvironmental permitEnvironmental permit. – Any permit, certificate, approval, or authorization issued by the Commission or the Department under this Article, Article 4 of Chapter 113A, Articles 9 or 10 of Chapter 130A, or Articles 21, 21A, or 21B of Chapter 143 of the General Statutes, or the rules adopted thereunder.G.S. § 143B-279.22(a)(4). Rules adopted by the CommissionCommissionCommission. – The Environmental Management Commission.G.S. § 143B-279.22(a)(2) pursuant to this section are not subject to G.S. 150B-19.4 or G.S. 150B-21.3(b3), and may do the following: (1) Define the permissible uses of artificial intelligence in the permitting process. (2) Establish quality assurance and control requirements for permits drafted with the assistance of artificial intelligence. (3) Adopt training requirements for agency staff who use artificial intelligence to review applications for environmental permitsEnvironmental permitEnvironmental permit. – Any permit, certificate, approval, or authorization issued by the Commission or the Department under this Article, Article 4 of Chapter 113A, Articles 9 or 10 of Chapter 130A, or Articles 21, 21A, or 21B of Chapter 143 of the General Statutes, or the rules adopted thereunder.G.S. § 143B-279.22(a)(4). (4) Establish confidentiality, cybersecurity, and data protection requirements for information processed, transmitted, stored, or generated through artificial intelligence systems, including requirements to protect confidential business information, trade secrets, security-sensitive infrastructure information, and other information protected from disclosure under State or federal law. (5) Prohibit the use of artificial intelligence systems that use applicant information, permit application materials, or agency records to train a vendor-owned or third-party artificial intelligence model unless expressly authorized by the DepartmentDepartmentDepartment. – The Department of Environmental Quality.G.S. § 143B-279.22(a)(3) under standards approved by the State Chief Information Officer. (6) Establish protocols for monitoring and discontinuing the use of artificial intelligence based on the performance and accuracy of artificial intelligence systems.
Subsection (c) grants the Environmental Management Commission discretionary authority to adopt rules governing six areas of AI use in permitting: permissible uses, quality assurance, staff training, data protection and confidentiality, restrictions on using agency data to train third-party AI models, and performance monitoring protocols. Notably, rules adopted under this section are exempt from the fiscal note and periodic review requirements of G.S. 150B-19.4 and G.S. 150B-21.3(b3). The provision authorizing rules to prohibit AI systems from using applicant data for third-party model training unless authorized under standards approved by the State Chief Information Officer is an unusual data governance provision specific to government AI procurement.
(d) 3 Agency Staff Review; Decision-making. – No environmental permitEnvironmental permitEnvironmental permit. – Any permit, certificate, approval, or authorization issued by the Commission or the Department under this Article, Article 4 of Chapter 113A, Articles 9 or 10 of Chapter 130A, or Articles 21, 21A, or 21B of Chapter 143 of the General Statutes, or the rules adopted thereunder.G.S. § 143B-279.22(a)(4) application shall be approved, denied, delayed, conditioned, or otherwise acted upon solely on the basis of artificial intelligence output. Relevant agency staff shall independently review, evaluate, and, as appropriate, modify any analysis, recommendation, draft permit, draft permit condition, deficiency letter, or other work product prepared with the assistance of artificial intelligence.
Subsection (d) is the bill's human-oversight mandate. It flatly prohibits any environmental permit application from being approved, denied, delayed, conditioned, or otherwise acted upon solely on the basis of AI output. Agency staff must independently review, evaluate, and where appropriate modify all AI-assisted work product — including analyses, recommendations, draft permits, draft permit conditions, and deficiency letters. This is a strong human-in-the-loop requirement that treats AI strictly as a decision-support tool, never as the decision-maker.
(e) 4 Transparency. – The DepartmentDepartmentDepartment. – The Department of Environmental Quality.G.S. § 143B-279.22(a)(3) shall maintain and make publicly available documents describing the artificial intelligence systems used by agency staff and the functions performed by those artificial intelligence systems in the environmental permitting process. The DepartmentDepartmentDepartment. – The Department of Environmental Quality.G.S. § 143B-279.22(a)(3) shall inform an applicant for an environmental permitEnvironmental permitEnvironmental permit. – Any permit, certificate, approval, or authorization issued by the Commission or the Department under this Article, Article 4 of Chapter 113A, Articles 9 or 10 of Chapter 130A, or Articles 21, 21A, or 21B of Chapter 143 of the General Statutes, or the rules adopted thereunder.G.S. § 143B-279.22(a)(4) to the fact that artificial intelligence was used by agency staff to assist in reviewing the permit application or to assist in drafting the permit.
Subsection (e) imposes two transparency obligations on the Department. First, it must maintain and make publicly available documents describing the AI systems used and their functions in the permitting process. Second, it must inform each permit applicant when AI was used to assist in reviewing their application or drafting their permit. The public-documentation requirement creates a standing disclosure obligation; the applicant-notification requirement creates a per-transaction disclosure triggered by AI involvement.
(f) Federal and State Law Compliance. – The CommissionCommissionCommission. – The Environmental Management Commission.G.S. § 143B-279.22(a)(2) and the DepartmentDepartmentDepartment. – The Department of Environmental Quality.G.S. § 143B-279.22(a)(3) shall implement the program established in this section only to the extent consistent with State law, federal law, and federal delegation agreements, federally approved State programs, and all applicable public notice, public comment, hearing, administrative review, and judicial review requirements. Nothing in this section shall be construed to alter the substantive standards applicable to the issuance, denial, modification, or revocation of an environmental permitEnvironmental permitEnvironmental permit. – Any permit, certificate, approval, or authorization issued by the Commission or the Department under this Article, Article 4 of Chapter 113A, Articles 9 or 10 of Chapter 130A, or Articles 21, 21A, or 21B of Chapter 143 of the General Statutes, or the rules adopted thereunder.G.S. § 143B-279.22(a)(4).
Subsection (f) is a savings clause requiring the program to operate within the bounds of existing state and federal law, including federal delegation agreements, approved state programs, and all applicable public-notice, comment, hearing, and review requirements. It expressly provides that the bill does not alter the substantive standards for permit issuance, denial, modification, or revocation. This provision creates no new compliance obligation.
(g)(1)–(5) 5 Reporting. – No later than January 15 each year, the DepartmentDepartmentDepartment. – The Department of Environmental Quality.G.S. § 143B-279.22(a)(3) shall submit a report to the Environmental Review CommissionCommissionCommission. – The Environmental Management Commission.G.S. § 143B-279.22(a)(2), the Joint Legislative Oversight Committee on Agriculture and Natural and Economic Resources, and the Fiscal Research Division. The report shall contain the following information from the previous year: (1) The number and type of permits for which artificial intelligence was used. (2) A performance assessment for each type of artificial intelligence system used. (3) A comparison of the average permit processing timeframes for agency staff using artificial intelligence relative to agency staff not using artificial intelligence. (4) A report on each incident in which agency staff have identified artificial intelligence system errors, and the corrective actions taken in response to these incidents. (5) Any recommendations as to whether to continue, expand, modify, or discontinue the use of artificial intelligence systems in the environmental permitting process.
Subsection (g) imposes an annual legislative reporting obligation on the Department. By January 15 each year, the Department must submit a report to the Environmental Review Commission, the Joint Legislative Oversight Committee on Agriculture and Natural and Economic Resources, and the Fiscal Research Division covering five categories: permit volumes where AI was used, performance assessments by AI system type, processing-time comparisons between AI-assisted and non-AI-assisted staff, incident reports identifying AI errors and corrective actions, and recommendations on whether to continue, expand, modify, or discontinue AI use. This is a structured operational performance reporting requirement.
(1)–(4) 6 Program Implementation. – The CommissionCommissionCommission. – The Environmental Management Commission.G.S. § 143B-279.22(a)(2) and the DepartmentDepartmentDepartment. – The Department of Environmental Quality.G.S. § 143B-279.22(a)(3) shall implement the Joint A.I. Environmental Permitting Program (Program) established in G.S. 143B-279.22, as enacted by Section 2 of this act, in the following sequence, with each phase to be initiated only upon the CommissionCommissionCommission. – The Environmental Management Commission.G.S. § 143B-279.22(a)(2) having determined that the applicable rules, quality assurance requirements, and staff training are sufficiently in place: (1) Phase I. – Post-construction stormwater permits issued under Article 21 of Chapter 143 of the General Statutes. (2) Phase II. – Major and minor new source review air permits issued under Article 21B of Chapter 143 of the General Statutes. (3) Phase III. – Erosion and sedimentation control permits issued under Article 4 of Chapter 113A of the General Statutes. (4) Phase IV. – Any other environmental permitEnvironmental permitEnvironmental permit. – Any permit, certificate, approval, or authorization issued by the Commission or the Department under this Article, Article 4 of Chapter 113A, Articles 9 or 10 of Chapter 130A, or Articles 21, 21A, or 21B of Chapter 143 of the General Statutes, or the rules adopted thereunder.G.S. § 143B-279.22(a)(4) to be determined by the CommissionCommissionCommission. – The Environmental Management Commission.G.S. § 143B-279.22(a)(2) following the Program's extension through the first three implementation phases as provided in this section.
Section 3 establishes a mandatory four-phase implementation sequence. Each phase covers a different permit category: Phase I covers post-construction stormwater permits, Phase II covers new source review air permits, Phase III covers erosion and sedimentation control permits, and Phase IV covers any additional permit types determined by the Commission. Each phase may be initiated only after the Commission determines that applicable rules, quality assurance requirements, and staff training are sufficiently in place. This is a government-internal operational sequencing requirement.
Staffing Flexibility. – Funds appropriated in this section may be used by the Director of the CommissionCommissionCommission. – The Environmental Management Commission.G.S. § 143B-279.22(a)(2)'s staff to hire temporary employees with demonstrated expertise in artificial intelligence technologies to provide technical guidance and assistance in implementing A.I. tools for purposes of improving the efficiency and quality of environmental permitEnvironmental permitEnvironmental permit. – Any permit, certificate, approval, or authorization issued by the Commission or the Department under this Article, Article 4 of Chapter 113A, Articles 9 or 10 of Chapter 130A, or Articles 21, 21A, or 21B of Chapter 143 of the General Statutes, or the rules adopted thereunder.G.S. § 143B-279.22(a)(4) processing under this act. Notwithstanding any other provision of law to the contrary, temporary employment or engagement under this section shall be funded from appropriations made available under this act and shall not be subject to the position classification or salary schedule requirements of Chapter 126 of the General Statutes.
Section 4 authorizes the Director of the Commission's staff to hire temporary employees with AI expertise to support program implementation. These temporary hires are exempt from the position classification and salary schedule requirements of Chapter 126 of the General Statutes, funded from appropriations under the act. This is an internal staffing provision that creates no external compliance obligation.
Appropriation. – There is appropriated from the General Fund to the Department of Environmental Quality the sum of one million dollars ($1,000,000) in nonrecurring funds for the 2026-2027 fiscal year to be used for purposes of implementing the Joint A.I. Environmental Permitting Program established in G.S. 143B-279.22, as enacted by Section 2 of this act. Funds appropriated under this section shall not revert at the end of the 2026-2027 fiscal year but shall remain available until expended.
Section 5 appropriates $1,000,000 in nonrecurring General Fund money to the Department of Environmental Quality for fiscal year 2026–2027 to implement the program. The funds do not revert at the end of the fiscal year but remain available until expended. This is a fiscal provision creating no compliance obligation.
Effective Date. – This act becomes effective July 1, 2026.
Section 6 sets the effective date of the act as July 1, 2026.