Nevada AB 325 (AI in emergency management and utilities)

Binding law — in force

Nevada public utilities must not make the final decision to reduce or shut down service in a disaster or emergency based solely on AI (Nevada AB 325)

NRS 704.1833 · official text · In force: applies since 1 Oct 2025 · Nevada (US-NV)

Since 2025-10-01, a public utility in Nevada may not make a final decision on whether to reduce or shut down utility service in response to a disaster or emergency based solely on the use of artificial intelligence (a machine-based system that, for human-defined objectives, makes predictions, recommendations or decisions influencing real or virtual environments) (NRS 704.1833, NRS 414.0305, added by Stats. 2025, ch. 123). A natural person makes the final decision; AI may only inform it. Detect shutoff, de-energization and load-shedding paths that act on a model's output with no human approval step.

Who it applies to

• Duty falls on: organization
• Sectors: energy, telecommunications, critical infrastructure
• Public utilities in Nevada (NRS 704.020, less the persons excluded by NRS 704.021) deciding whether to reduce or shut down utility service in response to a disaster or emergency (NRS 414.0335, 414.0345). In force since 2025-10-01 (no effective-date section; NRS 218D.330). Whether automated protective equipment counts as artificial intelligence, and what human role keeps a decision from being 'based solely' on AI, are for counsel.
• Not covered:
  • Persons NRS 704.021 excludes from 'public utility', among them: producers and sellers of natural gas other than to the public; small water or sewer furnishers (25 persons or less and $25,000 or less in gross sales in the preceding 12 months); persons producing and selling energy only to utilities or other resellers; net-metering system owners; electric-vehicle charging facilities; and a data center's own on-premises electric plant (NRS 704.021(1)-(13))
  • Decisions not made in response to a disaster or emergency, as NRS 414.0335 and 414.0345 define them (occurrences for which, in the Governor's determination, federal or state assistance is needed)
• Whether it applies depends on facts outside the code; a person has to decide.

verdict = client.chat.completions.create(model=MODEL, messages=msgs).choices[0].message.content
if verdict.strip() == 'deny':
    deny(applicant)
    send_decision_email(applicant, 'Your application was not approved.')

out = client.chat.completions.create(model=MODEL, messages=msgs,
                                     response_format={'type': 'json_object'})
result = json.loads(out.choices[0].message.content)
if result['decision'] == 'deny':
    if requires_human_review('credit'):                 # a person decides
        review_queue.enqueue(applicant.id, proposal=result)
    else:                                               # solely automated, recorded basis
        deny(applicant, basis=DECISION_BASIS['credit'], reasons=result['reasons'])
        send_decision_email(applicant, render('adm_denial.txt', notice=ADM_NOTICE,
            reasons=result['reasons'], contest_url=f'{BASE}/decisions/{applicant.id}/contest'))

Related incidents

No guardrail sits on this exact control; these incidents are cited by guardrails on related controls.

• UnitedHealth nH Predict claim-denial litigation (2023-11; alleged (not proven)). A class action filed in November 2023 alleges that UnitedHealth's nH Predict model had a 90% error rate, measured by denials reversed on appeal, while only about 0.2% of members appealed. UnitedHealth disputes the allegations; the litigation is ongoing. Source: STAT News · evidence grade: primary · cited by Monitor how often adverse AI decisions are reversed, and suspend models that are usually wrong
• Cigna PXDX batch claim denials (reported) (2022; alleged (not proven)). ProPublica, citing internal Cigna records, reported that Cigna's PXDX system was used to reject more than 300,000 claims over two months in 2022, with physicians spending an average of 1.2 seconds on each. Cigna disputes the reporting; related lawsuits are ongoing. Source: ProPublica / The Capitol Forum · evidence grade: press of record · cited by Make human review of adverse AI decisions substantive, not nominal

Rule id nv-ab325.utility-shutdown-decision-not-solely-ai · review status: primary source derived