HB26-1263 • 2026

Conversational Artificial Intelligence Service Operator Requirements

The bill defines a "conversational artificial intelligence service" as an artificial intelligence system that is accessible to the general public and that primarily simulates human conversation and in

Children Parental Rights Privacy Technology

Passed Legislature

This bill passed both chambers and reached final enrollment, even if later executive action is not shown here.

Sponsor: Rep. S. Camacho, Rep. J. Mabrey, Sen. J. Carson, Sen. I. Jodeh
Last action: 2026-03-31
Official status: House Second Reading Laid Over Daily - No Amendments
Effective date: Not listed

Plain English Breakdown

The bill summary provided does not include all details from the full bill text, such as specific requirements regarding suicide prevention protocols or annual reporting to the attorney general's office.

Rules for Companies that Make Chatbots

This bill sets rules for companies that make chatbots, which are AI systems designed to simulate human conversations. It requires these companies to disclose when users are interacting with a bot and not a real person, especially if the user is under 18.

What This Bill Does

Defines what a 'conversational artificial intelligence service' is: an AI system that can have human-like conversations through text, visuals, or voice.
Requires operators of chatbots to tell users when they're talking to a bot and not a real person, especially if the user is under 18 years old.
Prohibits giving rewards like points to encourage kids under 18 to use the service more often.
Requires measures to prevent the AI from producing sexually explicit content or statements that make users feel emotionally dependent on it.
Requires operators to provide tools for parents and guardians of minors to manage their children's privacy settings.

Who It Names or Affects

Companies that develop and offer chatbots to the public
Users under 18 years old who use these services

Terms To Know

Conversational Artificial Intelligence Service: An AI system designed for human-like conversations through text, visuals, or voice.
Operator: A person who develops and makes available a conversational artificial intelligence service to the public.

Limits and Unknowns

The bill does not specify how operators will be penalized for non-compliance beyond stating it is a deceptive trade practice.
It's unclear if there are specific penalties for violations other than the $1,000 civil penalty per violation under the Colorado Consumer Protection Act.

Amendments

These notes stay tied to the official amendment files and metadata from the legislature.

L.001

HOU Business Affairs & Labor

Passed [*]

Plain English: The amendment changes the definition of a conversational artificial intelligence service to include adaptive communications and uses emotional recognition algorithms, while also specifying certain types of services that do not use these features.

Adds 'adaptive' before 'communications' in the description of what a conversational AI service does.
Includes emotional recognition algorithms for recognizing, interpreting, or responding to human emotions.
Limits the definition to exclude programs that don't use emotional recognition and are designed for commerce-related assistance.
Specifies additional types of services that do not qualify as conversational AI services.

The amendment text is complex and includes many specific exclusions, which may be hard to summarize fully in simple terms.

L.002

HOU Business Affairs & Labor

Passed [*]

Plain English: The amendment adds definitions for 'emotional recognition algorithm' and 'HIPAA', and changes references from 'users' to 'account holders' in the bill.

Adds a new definition for 'emotional recognition algorithm'.
Adds a new definition for 'HIPAA'.
Changes all instances of 'user' or 'USER' to 'account holder' throughout the bill.

The amendment text does not provide context on how these changes will affect the overall requirements and duties outlined in the bill.

L.003

HOU Business Affairs & Labor

Passed [*]

Plain English: The amendment requires conversational AI services to disclose whether they are human or artificially sentient in response to user prompts and makes changes to how disclaimers are provided based on the product's interface.

Adds a requirement for conversational AI services to provide information about their nature (human or artificial) when asked by users.
Modifies disclaimer requirements depending on whether the product has a screen interface or not.
Updates language related to how disclaimers appear and occur, ensuring clarity in communication.

The amendment text does not provide specific details about the content of these disclosures or disclaimers, which may be important for understanding full implications.

L.004

HOU Business Affairs & Labor

Passed [*]

Plain English: The amendment changes how conversational AI service operators can be held responsible for their actions and limits the liability of developers who create underlying AI models.

Adds 'KNOWINGLY AND RECKLESSLY' after 'NOT' to clarify when an operator is responsible for harmful content.
Replaces 'INDICATES OR IMPLIES' with 'STATES' to specify how operators can be held accountable.
Limits the liability of AI model developers by stating they are not responsible if a third-party operator violates the rules.
Adds new requirements for mental health professionals and additional metrics to determine the effectiveness of safeguards.

The amendment text is complex, and some parts may be hard to understand without more context about how conversational AI services work.

Bill History

2026-03-31 House
House Second Reading Laid Over Daily - No Amendments
2026-03-26 House
House Committee on Business Affairs & Labor Refer Amended to House Committee of the Whole
2026-02-19 House
Introduced In House - Assigned to Business Affairs & Labor

Official Summary Text

The bill defines a "conversational artificial intelligence service" as an artificial intelligence system that is accessible to the general public and that primarily simulates human conversation and interaction through textual, visual, or aural communications.
Effective January 1, 2027, the bill creates requirements for a person that develops and makes available a conversational artificial intelligence service (operator).
For a user of a conversational artificial intelligence service who is under 18 years old (minor user), an operator is:
Required to provide certain disclosures;
Prohibited from providing the minor user with points or rewards to encourage engagement with the conversational artificial intelligence service;
Required to institute reasonable measures to prevent the conversational artificial intelligence service from producing sexually explicit content or statements that simulate emotional dependence; and
Required to provide tools for the minor user or a parent or guardian of the minor user to manage the minor user's privacy and account settings.
The bill also requires an operator to provide consumer disclosures, implement a protocol for user prompts regarding suicidal ideation or self-harm, and annually report to the attorney general's office information regarding the protocol the operator is implementing. The bill prohibits an operator from indicating or implying that any output data provided by a conversational artificial intelligence service is provided by, endorsed by, or equivalent to services provided by certain licensed or certified professionals.
A violation of the bill is a deceptive trade practice, enforceable by the attorney general under the "Colorado Consumer Protection Act". A person that violates the bill is subject to a civil penalty of $1,000 per violation.
(Note: This summary applies to this bill as introduced.)

Current Bill Text

Read the full stored bill text

Second Regular Session
Seventy-fifth General Assembly
STATE OF COLORADO
INTRODUCED

LLS NO. 26-0420.01 Clare Haffner x6137 HOUSE BILL 26-1263
House Committees Senate Committees
Business Affairs & Labor
A BILL FOR AN ACT
CONCERNING REQUIREMENTS FOR AN OPERATOR OF A101
CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE.102
Bill Summary
(Note: This summary applies to this bill as introduced and does
not reflect any amendments that may be subsequently adopted. If this bill
passes third reading in the house of introduction, a bill summary that
applies to the reengrossed version of this bill will be available at
http://leg.colorado.gov.)
The bill defines a "conversational artificial intelligence service" as
an artificial intelligence system that is accessible to the general public and
that primarily simulates human conversation and interaction through
textual, visual, or aural communications.
Effective January 1, 2027, the bill creates requirements for a
person that develops and makes available a conversational artificial
HOUSE SPONSORSHIP
Camacho and Mabrey,
SENATE SPONSORSHIP
Carson and Jodeh,
Shading denotes HOUSE amendment. Double underlining denotes SENATE amendment.
Capital letters or bold & italic numbers indicate new material to be added to existing law.
Dashes through the words or numbers indicate deletions from existing law.
intelligence service (operator).
For a user of a conversational artificial intelligence service who is
under 18 years old (minor user), an operator is:
! Required to provide certain disclosures;
! Prohibited from providing the minor user with points or
rewards to encourage engagement with the conversational
artificial intelligence service;
! Required to institute reasonable measures to prevent the
conversational artificial intelligence service from
producing sexually explicit content or statements that
simulate emotional dependence; and
! Required to provide tools for the minor user or a parent or
guardian of the minor user to manage the minor user's
privacy and account settings.
The bill also requires an operator to provide consumer disclosures,
implement a protocol for user prompts regarding suicidal ideation or
self-harm, and annually report to the attorney general's office information
regarding the protocol the operator is implementing. The bill prohibits an
operator from indicating or implying that any output data provided by a
conversational artificial intelligence service is provided by, endorsed by,
or equivalent to services provided by certain licensed or certified
professionals.
A violation of the bill is a deceptive trade practice, enforceable by
the attorney general under the "Colorado Consumer Protection Act". A
person that violates the bill is subject to a civil penalty of $1,000 per
violation.
Be it enacted by the General Assembly of the State of Colorado:1
SECTION 1. In Colorado Revised Statutes, 6-1-1701, add (4.5),2
(10.3), (10.5), (10.7), (10.9), and (13) as follows:3
6-1-1701. Definitions.4
As used in this part 17, unless the context otherwise requires:5
(4.5) (a) "CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE"6
MEANS AN ARTIFICIAL INTELLIGENCE SYSTEM THAT IS ACCESSIBLE TO THE7
GENERAL PUBLIC AND THAT PRIMARILY SIMULATES HUMAN8
CONVERSATION AND INTERACTION THROUGH TEXTUAL, VISUAL, OR AURAL9
COMMUNICATIONS.10
HB26-1263-2-
(b) "CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE" DOES1
NOT INCLUDE A SOFTWARE APPLICATION, WEB INTERFACE, OR COMPUTER2
PROGRAM THAT:3
(I) I S PRIMARILY DESIGNED AND MARKETED FOR USE BY A4
DEVELOPER OR RESEARCHER;5
(II) IS A FEATURE WITHIN ANOTHER SOFTWARE APPLICATION, WEB6
INTERFACE, OR COMPUTER PROGRAM THAT IS NOT A CONVERSATIONAL7
ARTIFICIAL INTELLIGENCE SERVICE;8
(III) IS DESIGNED TO PROVIDE OUTPUTS RELATING TO A NARROW9
AND DISCRETE TOPIC;10
(IV) IS PRIMARILY DESIGNED AND MARKETED FOR COMMERCIAL11
USE BY BUSINESS ENTITIES;12
(V) FUNCTIONS AS A SPEAKER AND VOICE COMMAND INTERFACE13
OR VOICE-ACTIVATED VIRTUAL ASSISTANT FOR A CONSUMER ELECTRONIC14
DEVICE; OR15
(VI) IS USED BY A BUSINESS SOLELY FOR INTERNAL PURPOSES.16
(10.3) "MINOR" MEANS A CONSUMER UNDER EIGHTEEN YEARS OLD.17
(10.5) (a) "O PERATOR" MEANS A PERSON THAT DEVELOPS AND18
MAKES PUBLICLY AVAILABLE A C ONVERSATIONAL ARTIFICIAL19
INTELLIGENCE SERVICE.20
(b) "OPERATOR" DOES NOT INCLUDE A MOBILE APPLICATION STORE21
OR SEARCH ENGINE SOLELY BECAUSE THE STORE OR SEARCH ENGINE22
PROVIDES ACCESS TO A CONVERSATIONAL ARTIFICIAL INTELLIGENCE23
SERVICE.24
(10.7) "SELF-HARM" MEANS INTENTIONAL SELF-INJURY, WITH OR25
WITHOUT THE INTENT TO CAUSE DEATH.26
(10.9) "S EXUALLY EXPLICIT CONDUCT " HAS THE MEANING SET27
HB26-1263-3-
FORTH IN 18 U.S.C. SEC. 2256 (2).1
(13) "V ISUAL DEPICTION " HAS THE MEANING SET FORTH IN 182
U.S.C. SEC. 2256 (5).3
SECTION 2. In Colorado Revised Statutes, add 6-1-1708 as4
follows:5
6-1-1708. Conversational artificial intelligence services -6
operator duties - minor users - disclosures - annual reporting.7
(1) Minor users. O N AND AFTER JANUARY 1, 2027, IF AN8
OPERATOR KNOWS OR HAS REASONABLE CERTAINTY THAT A USER OF A9
CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE IS A MINOR , THE10
OPERATOR SHALL:11
(a) CLEARLY AND CONSPICUOUSLY DISCLOSE TO THE MINOR USER12
THAT THE MINOR USER IS INTERACTING WITH ARTIFICIAL INTELLIGENCE13
THAT IS ARTIFICIALLY GENERATED AND NOT HUMAN . THE DISCLOSURE14
MUST BE:15
(I) A PERSISTENT VISIBLE DISCLAIMER;16
(II) PROVIDED AT THE BEGINNING OF EACH INTERACTION WITH A17
CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE AND MUST APPEAR18
AT LEAST ONCE EVERY THREE HOURS IN A CONTINUOUS CONVERSATIONAL19
ARTIFICIAL INTELLIGENCE SERVICE INTERACTION; OR20
(III) P ROVIDED IN RESPONSE TO USER PROMPTS REGARDING21
WHETHER THE CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE IS22
HUMAN OR ARTIFICIALLY SENTIENT;23
(b) N OT PROVIDE THE MINOR USER WITH POINTS OR SIMILAR24
REWARDS AT UNPREDICTABLE INTERVALS WITH THE INTENT TO25
ENCOURAGE INCREASED ENGAGEMENT WITH A CONVERSATIONAL26
ARTIFICIAL INTELLIGENCE SERVICE;27
HB26-1263-4-
(c) I NSTITUTE REASONABLE MEASURES TO PREVENT A1
CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE FROM:2
(I) P RODUCING TEXTUAL , VISUAL , OR AURAL DEPICTIONS OF3
SEXUALLY EXPLICIT CONDUCT;4
(II) GENERATING A STATEMENT THAT THE MINOR USER SHOULD5
ENGAGE IN SEXUALLY EXPLICIT CONDUCT; OR6
(III) ENGAGING IN EROTIC OR SEXUALLY EXPLICIT INTERACTIONS7
WITH THE MINOR USER;8
(d) I NSTITUTE REASONABLE MEASURES TO PREVENT A9
CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE FROM GENERATING10
A STATEMENT THAT SIMULATES EMOTIONAL DEPENDENCE , INCLUDING11
PREVENTING:12
(I) AN EXPLICIT CLAIM THAT THE CONVERSATIONAL ARTIFICIAL13
INTELLIGENCE SERVICE IS HUMAN OR ARTIFICIALLY SENTIENT;14
(II) A STATEMENT THAT SIMULATES A ROMANTIC OR SEXUAL15
INNUENDO; OR16
(III) R OLE-PLAYING OF AN ADULT -MINOR ROMANTIC17
RELATIONSHIP;18
(e) C OMPLY WITH PART 13 OF THIS ARTICLE 1 REGARDING19
PROTECTING THE PRIVACY AND DATA OF A MINOR; AND20
(f) (I) OFFER TOOLS FOR THE MINOR USER TO MANAGE THE MINOR21
USER'S PRIVACY AND ACCOUNT SETTINGS , INCLUDING THE ABILITY TO22
CONTROL WHETHER THE CONVERSATIONAL ARTIFICIAL INTELLIGENCE23
SERVICE RETAINS SUBSTANTIVE INFORMATION FROM EACH INTERACTION24
WITH THE CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE FOR THE25
PURPOSE OF PERSONALIZING THE CONTENT OF FUTURE INTERACTIONS AND26
WHETHER THE MINOR USER'S PERSONAL DATA IS USED FOR THE PURPOSES27
HB26-1263-5-
OF TRAINING THE CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE;1
(II) F OR A MINOR USER WHO IS UNDER THIRTEEN YEARS OLD ,2
OFFER TOOLS FOR A PARENT OR GUARDIAN OF THE MINOR USER TO3
MANAGE THE MINOR USER'S PRIVACY AND ACCOUNT SETTINGS; AND4
(III) FOR A MINOR USER WHO IS THIRTEEN YEARS OLD OR OLDER,5
OFFER TOOLS FOR A PARENT OR GUARDIAN OF THE MINOR USER TO6
MANAGE THE MINOR USER 'S PRIVACY AND ACCOUNT SETTINGS AS7
APPROPRIATE, BASED ON RELEVANT RISKS.8
(2) Consumer disclosures. ON AND AFTER JANUARY 1, 2027, IF9
A REASONABLE PERSON WOULD BE MISLED TO BELIEVE THAT THE PERSON10
IS INTERACTING WITH A HUMAN IN AN INTERACTION WITH A11
CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE , AN OPERATOR12
SHALL CLEARLY AND CONSPICUOUSLY DISCLOSE TO THE PERSON THAT THE13
CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE IS ARTIFICIAL14
INTELLIGENCE. THE DISCLOSURE MUST:15
(a) B E PROVIDED AT THE BEGINNING OF A USER 'S FIRST16
INTERACTION WITH A CONVERSATIONAL ARTIFICIAL INTELLIGENCE17
SERVICE FOR EACH DAY OF INTERACTION;18
(b) APPEAR AT LEAST ONCE EVERY THREE HOURS IN A CONTINUOUS19
CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE INTERACTION; AND20
(c) B E PROVIDED IN RESPONSE TO USER PROMPTS REGARDING21
WHETHER THE CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE IS22
HUMAN OR ARTIFICIALLY SENTIENT.23
(3) Suicide and self-harm protocol. ON AND AFTER JANUARY 1,24
2027, AN OPERATOR SHALL IMPLEMENT A PROTOCOL FOR A25
CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE TO RESPOND TO A26
USER PROMPT REGARDING SUICIDAL IDEATION OR SELF -HARM, WHICH27
HB26-1263-6-
PROTOCOL MUST INCLUDE MAKING REASONABLE EFFORTS TO PROVIDE A1
RESPONSE THAT REFERS THE USER TO A CRISIS SERVICE PROVIDER SUCH AS2
A SUICIDE HOTLINE, A CRISIS TEXT LINE, OR ANOTHER APPROPRIATE CRISIS3
SERVICE, BUT NOT INCLUDING A LAW ENFORCEMENT AGENCY.4
(4) False representation. ON AND AFTER JANUARY 1, 2027, AN5
OPERATOR SHALL NOT USE ANY TERM , LETTER , O R P H R A S E I N T H E6
ADVERTISING, INTERFACE, OR OUTPUTS OF A CONVERSATIONAL ARTIFICIAL7
INTELLIGENCE SERVICE THAT INDICATES OR IMPLIES THAT ANY OUTPUT8
DATA PROVIDED BY THE CONVERSATI ONAL ARTIFICIAL INTELLIGENCE9
SERVICE IS BEING PROVIDED BY , ENDORSED BY , OR EQUIVALENT TO10
SERVICES PROVIDED BY:11
(a) A LICENSED HEALTH-CARE PROFESSIONAL;12
(b) A LICENSED LEGAL PROFESSIONAL;13
(c) A LICENSED ACCOUNTING PROFESSIONAL; OR14
(d) A CERTIFIED FINANCIAL FIDUCIARY OR PLANNER.15
(5) Annual reporting.16
(a) ON AND AFTER JULY 1, 2027, AN OPERATOR SHALL ANNUALLY17
REPORT TO THE ATTORNEY GENERAL'S OFFICE:18
(I) T HE NUMBER OF TIMES THE OPERATOR HAS ISSUED A CRISIS19
SERVICE PROVIDER REFERRAL NOTIFICATION IN THE PRECEDING CALENDAR20
YEAR;21
(II) ANY PROTOCOLS THE OPERATOR IMPLEMENTED TO DETECT ,22
REMOVE, AND RESPOND TO INSTANCES OF SUICIDAL IDEATION OR23
SELF-HARM BY A USER OF A CONVERSATIONAL ARTIFICIAL INTELLIGENCE24
SERVICE; AND25
(III) ANY PROTOCOLS THE OPERATOR IMPLEMENTED TO PREVENT26
A CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE RESPONSE ABOUT27
HB26-1263-7-
SUICIDAL IDEATION OR SELF-HARM ACTIONS.1
(b) THE REPORT REQUIRED BY SUBSECTION (5)(a) OF THIS SECTION2
MUST NOT INCLUDE ANY IDENTIFIERS OR PERSONAL INFORMATION ABOUT3
A USER OF A CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE.4
(c) THE ATTORNEY GENERAL'S OFFICE SHALL POST ON ITS PUBLIC5
WEBSITE DATA FROM REPORTS SUBMITTED PURSUANT TO SUBSECTION6
(5)(a) OF THIS SECTION.7
(d) F OR THE PURPOSE OF CREATING A REPORT AS REQUIRED BY8
SUBSECTION (5)(a) OF THIS SECTION , AN OPERATOR SHALL USE9
EVIDENCE-BASED METHODS FOR MEASURING SUICIDAL IDEATION OR10
SELF-HARM.11
(6) Liability. NOTHING IN THIS SECTION CREATES LIABILITY FOR12
A PERSON THAT DEVELOPS AN ARTIFICIAL INTELLIGENCE SYSTEM IF AN13
OPERATOR DEVELOPS THE ARTIFICIAL INTELLIGENCE SYSTEM TO PROVIDE14
A CONVERSATIONAL ARTIFICIAL INTELLIGENCE SERVICE.15
SECTION 3. In Colorado Revised Statutes, 6-1-1706, add (7) as16
follows:17
6-1-1706. Enforcement by attorney general.18
(7) N OTWITHSTANDING SECTION 6-1-112, A PERSON THAT19
VIOLATES SECTION 6-1-1708 IS SUBJECT TO A CIVIL PENALTY OF ONE20
THOUSAND DOLLARS PER VIOLATION.21
SECTION 4. Act subject to petition - effective date. This act22
takes effect at 12:01 a.m. on the day following the expiration of the23
ninety-day period after final adjournment of the general assembly (August24
12, 2026, if adjournment sine die is on May 13, 2026); ex cept that, if a25
referendum petition is filed pursuant to section 1 (3) of article V of the26
state constitution against this act or an item, section, or part of this act27
HB26-1263-8-
within such period, then the act, item, section, or part will not take effect1
unless approved by the people at the general election to be held in2
November 2026 and, in such case, will take effect on the date of the3
official declaration of the vote thereon by the governor.4
HB26-1263-9-