Back to the agents

US AI rules, US AI rules

Colo. Rev. Stat. § 6-1-1701 et seq.

Jurisdiction

State CO, State IL, State UT

Risk dimensions

operations, employment, privacy

Tier

Tier 1

Coverage

55 %

Impact

5/5

Read

2026-08-15

Source chain

Every block in the register row has its own address, so an outcome can be cited down to the paragraph. The node carries the reference, the scope, the link to the official text and the outcomes in the tree that rest on the block.

• Developer duties

  Colo. Rev. Stat. § 6-1-1702

  Reasonable care against algorithmic discrimination, plus documentation handed to the deployer.

• Deployer duties

  Colo. Rev. Stat. § 6-1-1703

  A risk management program, an impact assessment and notice to the consumer before a consequential decision.

• Impact assessment

  Colo. Rev. Stat. § 6-1-1703(3)

  Purpose, known risks, data categories, performance metrics and post-deployment monitoring.

• Consumer notice and appeal

  Colo. Rev. Stat. § 6-1-1703(4)

  Notice of an adverse consequential decision, the reason, and a route to correction and human review.

• Public statement

  Colo. Rev. Stat. § 6-1-1703(5)

  A public summary of the high-risk systems deployed and how risk is managed.

• Discriminatory use

  775 ILCS 5/2-102(L)

  Use of artificial intelligence that has a discriminatory effect on a protected class in recruitment, promotion, discipline or discharge.

• Zip code as proxy

  775 ILCS 5/2-102(L)(2)

  Use of zip code as a proxy for a protected class.

• No defence for the system

  Utah Code § 13-72-101

  A person cannot rely on generative AI as a defence for a statement that breaches consumer protection law.

• Disclosure on request

  Utah Code § 13-72-201

  A clear statement that the counterpart is a generative AI system when the consumer asks.

• Regulated occupations

  Utah Code § 13-72-202

  Prominent disclosure at the start of the interaction in occupations that require a licence.

Colo. Rev. Stat. § 6-1-1701 et seq.

Rule tree

Rules are tested top down. Conditions are statutory elements and each outcome points to the register blocks it rests on.

1. ai-co-deployer

   Impact assessment missing for a high-risk system in Colorado

   prohibited · Requirement applies · deployer, impact, notice

   Colo. Rev. Stat. § 6-1-1703 requires a risk management program, an impact assessment and notice to the consumer before a consequential decision, with correction and human review.

2. ai-co

   The Colorado AI Act applies to the system

   risk · Requirement applies · developer, deployer, public

   Developers must use reasonable care against algorithmic discrimination and provide documentation (§ 6-1-1702). Deployers need a risk program, impact assessment, consumer notice and a public statement (§ 6-1-1703).

3. ai-il

   Illinois rules on AI in employment apply

   risk · Requirement applies

   775 ILCS 5/2-102(L) prohibits AI that discriminates in employment decisions and requires notice to the applicant. 820 ILCS 42 requires notice and consent for AI analysis of video interviews.

4. ai-ut

   Utah requires disclosure of generative AI

   risk · Requirement applies

   Utah Code § 13-75-103 requires disclosure on request that the consumer is interacting with generative AI, and regulated occupations must disclose it up front.

Outcome

Monitoring

The journal shows what moved in the register row, with the hash before and after. The impact score follows a rule stated in plain words.

How impact is scored: Grund: tillagd eller borttagen uppgift ger 3, ändrad uppgift 2, omläsning utan ändring 0. Tillägg: +1 när ändringen rör status eller tillämpningsdatum. Tillägg: +1 när agenten ligger i klass 1. Siffran begränsas till 0 till 5.

• 2026-08-15 · 0/5 · lasning

  Raden läst mot den officiella publiceringen utan ändring

• 2026-08-15 · 0/5 · lasning

  Raden läst mot den officiella publiceringen utan ändring

• 2026-08-15 · 0/5 · lasning

  Raden läst mot den officiella publiceringen utan ändring

• 2026-06-30 · 4/5 · tillampningsdatum, status

  Regeln började tillämpas enligt utgivaren

• 2026-01-01 · 4/5 · tillampningsdatum, status

  Regeln började tillämpas enligt utgivaren

• 2024-08-09 · 5/5 · antagande, identifierare, status

  Regeln antogs enligt 775 ILCS 5/2-102

• 2024-05-17 · 5/5 · antagande, identifierare, status

  Regeln antogs enligt Colo. Rev. Stat. § 6-1-1701 et seq.

• 2024-05-01 · 4/5 · tillampningsdatum, status

  Regeln började tillämpas enligt utgivaren

• 2024-03-13 · 5/5 · antagande, identifierare, status

  Regeln antogs enligt Utah Code § 13-72

Supervision

• Colorado Department of Law, Attorney General
• Illinois Department of Human Rights
• Utah Division of Consumer Protection

European counterparts

• ai-act · Båda texterna bygger på riskklasser, konsekvensbedömning och transparens. Coloradolagen är skriven kring beslut med rättslig verkan för konsumenter, inte kring att släppa ut en produkt på marknaden.
• ai-act · Anställning är ett högriskområde i bilaga III till AI-förordningen. Illinoisvägen går genom medborgarrättslagstiftning i stället för produktregler.
• ai-act · Artikel 50 i AI-förordningen bär samma tanke om transparens mot den som samspelar med systemet.

Artifacts and integrity

• https://legal.exploreworldai.com/api/public/v1/us-agents/us-ai-rules/manifest.json
• https://legal.exploreworldai.com/api/public/v1/us-agents/us-ai-rules/run
• https://legal.exploreworldai.com/api/public/v1/us-agents/us-ai-rules/monitor.json
• sha256:16018990250bcc2d3503fed2d00e7e4a9555e5f632f79d03e3bb14e560a16112
• sha256:498e73e8201acb70db77a0555cfa9d9759136c226813e51a7105a6b63545a35e

The verdict is a machine classification of the outcome, not legal advice and not a compliance decision. Responsibility position · Colo. Rev. Stat. § 6-1-1701 et seq.