Summary of HB 1161 (Session 2025, North Carolina) — Omnibus Artificial Intelligence Protections
Purpose
- Establishes a broad set of protections governing the deployment and use of artificial intelligence (AI) and related technologies across elections, education, employment, the courts, and the insurance industry in North Carolina.
- Includes specific prohibitions, policy authorizations for schools, a comprehensive framework for AI hiring governance ( dubbed the North Carolina Fair AI Hiring Act, NC-FAHA), standards for classroom AI use, insurance claim processing rules, court-related AI safeguards, and targeted funding for implementation.
Key Provisions by Part
Part I: Elections
- Section 163-278.18A — Prohibit AI in political advertisements
- Prohibits any use of AI in political ads (images, videos, voices, writings) by candidates, campaigns, parties, PACs, referendums, or sponsors.
- Violations: Class 1 misdemeanor; fines equal to 10 times the production/distribution cost of the AI-generated ad.
Penalties and sanctions:
- Candidate campaigns: cannot solicit donations or make contributions for 2 years following conviction.
- PACs: bank accounts frozen; cannot solicit donations or contribute; chair/director barred from soliciting, contributing, or working with other PACs for 5 years.
- Other individuals misusing AI to impersonate a candidate with intent to confuse/interfere: fraud/election interference charges (with exceptions for clearly labeled AI or parody).
- Effective date: Applies to AI-enabled political ads on or after the law’s effective date.
Part II: Education
- Section 2.1(a–f) — Local policies on student use of AI in schools
- Local boards may adopt AI-use policies (e.g., blocking AI on devices or networks; disciplinary actions for unauthorized use, including potential grade retention for repeated violations).
- Exceptions for teacher-authorized use or required by an IEP/504 plan; parents and principal must be notified if an exception is used.
- School districts and charter schools’ policies must align with new standards in 115C-102.13.
- Section 2.2(a–f) — Middle school curriculum and CS instruction
- Mandates a middle school course on critical thinking and civics, including how to identify disinformation and how government works (with references to classical works such as Machiavelli and Plato).
- Requires an introductory CS course for middle schoolers with topics including AI usage in academics, identifying AI-generated content, and cybersecurity.
- A $500,000 nonrecurring appropriation in 2026-2027 to implement these educational provisions.
- Effective beginning with the 2026-2027 school year.
- Section 2.3 — Higher education standards for AI in classrooms
- UNC system and community colleges must develop standards for AI use in classrooms, with annual student notification and disciplinary actions for misuse (e.g., failing grade).
- Effective for the 2026-2027 academic year.
Part III: Employment
- Chapter 95A — Fair Artificial Intelligence Hiring Act (NC-FAHA)
- Effective October 1, 2026.
Foundational aims:
- Ensure AEDTs (Automated Employment Decision Tools) used in hiring/promotions undergo independent bias audits; reveal results publicly; provide advance notice to affected applicants.
- Definitions (Section 95A-2):
- AEDT: any AI/ML-based tool that substantially assists or replaces discretionary judgment in a covered employment decision.
- Bias audit: independent, comprehensive audit assessing rates by race/ethnicity/sex, impact ratios, data source, and adverse impact.
- Other defined terms: Deployer (employer using AEDT), Independent auditor, Public/private employers, etc.
- Bias audit requirements (Section 95A-3)
- AEDTs must be audited within the prior 12 months; audit results publicly accessible.
- Disclosure and advance notice (Sections 95A-4 and 95A-5)
- Public summaries of bias audits must be posted within 30 days of audit completion.
- Applicants/employees must receive at least 10 business days’ notice before AEDT use, with the right to request an alternative process.
- Enforcement and penalties (Sections 95A-6 to 95A-7)
- AG can sue private employers; penalties scale per day for violations.
- Civil remedies include damages, injunctions, attorneys’ fees, etc.
- SHRC enforcement for public employers; potential withholdings against agency budgets to enforce compliance.
- Standards and government contracting (Sections 95A-8 to 95A-9)
- SHRC to publish standards for auditors, bias audits, and notices.
- Contractors with state contracts must certify AEDT use, disclose bias audits, and may be subject to contract breaches if noncompliant.
- Exemptions (Section 95A-10)
- Tools for standardized tests, small employers (fewer than 15 employees), certain recruitment tools, certain background checks, or federal compliance tools are exempt.
- Relationship to other laws (Section 95A-11)
- Supplements existing anti-discrimination laws; does not preempt local ordinances; severability clause.
Part IV: Insurance
- Section 4.1 — Limit AI use in insurance claim processing
- Adds to unfair claim settlement practices: using AI as the primary method to process a claim can be unfair.
- Effective October 1, 2026, applying to policies issued or renewed on/after that date.
Part V: Courts
- Article 64 (Chapter 7A)
- Defines AI terms (AI, deepfake, generative AI, hallucination) and creates judicial safeguards:
- Judges may dismiss cases involving any use of generative AI in court filings or appearances (without prejudice).
- Remedies for improper use may include new action, appeal, and restrictions on second use of AI (with potential sanctions and possible referral to the bar for attorneys).
- Effective July 1, 2026 for court documents filed and proceedings initiated on/after that date.
Part VI: Appropriation
- Allocates funds for implementation
- SHRC/OSHR/DOJ/Department of Administration: combined appropriations totaling roughly $700,000+ for 2026-2027 to support rulemaking, audits, oversight, and implementation.
Part VII: Effective Date
- General effective date: upon becoming law, with specific provisions triggering at different times noted above (e.g., 2026-2027 school year, October 1, 2026, July 1, 2026).
Potential Impacts
- Elections: Strong deterrence against AI-generated political ads; potential chilling effect on political campaigning and advertising.
- Education: Expanded AI governance in schools; mandatory civics/critical thinking content; increased funding to implement new policies and courses.
- Employment: Comprehensive regulatory framework for AEDTs, emphasizing transparency, bias auditing, and worker protections; enhanced enforcement mechanisms for both private and public employers.
- Insurance: Higher scrutiny of AI-driven claim processing to prevent unfair practices.
- Courts: Clear prohibitions and remedial pathways for AI usage in litigation and filings.
- Public agencies: Increased oversight and annual reporting on AEDT deployments.
Implementation and Funding
- Nonrecurring and ongoing funding provided to key state agencies to execute the Act’s provisions.
- Agencies tasked with rulemaking, auditing, enforcement, and education standards will play central roles.
Notes
- Some provisions include explicit exemptions and local flexibility, particularly in education and contracting.
- The bill creates broad, cross-cutting governance of AI use across several sectors, with substantial compliance and enforcement mechanisms.
