Technology (Artificial Intelligence)

In 2024, the general assembly enacted Senate Bill 24-205, which created consumer protections in interactions with artificial intelligence systems. The act repeals and reenacts those provisions with new requirements regarding the use of automated decision-making technology in consequential decisions.     The act defines an 'automated decision-making technology' (ADMT) as a technology that processes personal data and uses computation to generate output, including predictions, recommendations, classifications, rankings, scores, or other information that is used to make, guide, or assist a decision, judgment, or determination concerning an individual. The act defines a 'consequential decision' as a decision that relates to an individual's access to, eligibility for, or compensation related to education, employment, housing, financial or lending services, insurance, health-care services, or essential government services and public benefits.     The act requires the developer of an ADMT (developer) that is used to materially influence a consequential decision (covered ADMT), starting January 1, 2027, to provide a deployer of a covered ADMT (deployer) with technical documentation describing the covered ADMT's intended uses, categories of training data, known limitations, and instructions for appropriate use and human review. Developers must notify deployers of material updates or modifications to the covered ADMT. Both developers and deployers are required to retain records necessary to demonstrate compliance with the act for at least 3 years.     The act establishes consumer notice requirements, mandating that deployers provide clear and conspicuous notice to consumers at the point of interaction with a covered ADMT. A deployer is required to provide a consumer with a plain language description of a covered ADMT's role within 30 days after the covered ADMT makes a consequential decision that results in an adverse outcome for the consumer. The attorney general must adopt rules to clarify these post-adverse outcome disclosure requirements by January 1, 2027.     Consumers have the right to request personal data and correction of factually incorrect personal data used by a covered ADMT. The act also grants consumers the right to request meaningful human review and reconsideration following a covered ADMT making a consequential decision resulting in an adverse outcome.     The attorney general is directed to enforce the act through the 'Colorado Consumer Protection Act', and a violation of the act is deemed a deceptive trade practice. Before initiating an action before January 1, 2030, the attorney general must provide the developer or deployer with a 60-day notice and opportunity to cure the alleged violation, if a cure is deemed possible. The act does not create a new private right of action but establishes how fault is allocated between developers and deployers in civil actions alleging unlawful discrimination under existing law.     Specified entities are exempted from the requirements of the act to the extent the entities comply with other legal obligations.(Note: This summary applies to this bill as enacted.)

The act 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 adaptive textual, visual, or aural communications.     Effective January 1, 2027, the act creates requirements and prohibitions for a person, partnership, corporation, or entity that develops and makes publicly available a conversational artificial intelligence service or offers a conversational artificial intelligence service to a consumer (operator).     An operator is required to use commercially reasonable methods or generally accepted methods to estimate the age of a consumer who has or opens an account or profile to use a conversational artificial intelligence service (account holder) and the age of other users of a conversational artificial intelligence service. If an operator knows that an account holder or user is a minor, an operator is:Required to provide certain disclosures;Prohibited from providing the minor account holder or minor user with points or rewards to encourage engagement with the conversational artificial intelligence service;Required to institute technically feasible measures to prevent the conversational artificial intelligence service from producing explicit sexual conduct, intimate digital depictions, or statements that simulate emotional dependence;Required to implement a protocol for a conversational artificial intelligence service to stop engaging in response to a user prompt regarding sexual conduct with a minor; andRequired to provide tools for the minor account holder or minor user or a parent or guardian of the minor account holder or minor user to manage the minor account holder's or minor user's privacy and account settings.     The act also requires an operator to provide a disclosure to a user that a conversational artificial intelligence service is artificial intelligence, 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 act prohibits an operator from stating 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.     The act clarifies that nothing in the act limits an individual's ability to access certain information and resources pursuant to the state constitution, requires an operator to disclose confidential information, or authorizes content moderation practices inconsistent with the United States constitution.(Note: This summary applies to this bill as enacted.)

Surveillance data is defined in the act as data that is obtained through observation, inference, or surveillance of consumers or workers and that is related to personal characteristics, online behaviors, or biometrics of an individual or group, band, class, or tier to which the individual belongs. The definition of 'worker' in the act excludes federal and state employees and employees of public entities.     The act prohibits discrimination against a consumer or worker resulting from the use of a price or wage setting algorithm (PWSA) that uses statistical modeling, data analytics, artificial intelligence, or other data processing techniques to analyze surveillance data, the output of which is a substantial factor in:Individualized price setting used to determine the amount charged to a consumer; orIndividualized wage setting used to determine the wage offered to a worker.     The act specifies activities that are not individualized price or wage setting, as well as exemptions from the prohibition on price or wage setting. A person has not engaged in individualized price setting if the person can demonstrate, as described in the act, that differential prices are:Based on differences in the cost in providing a good or service to different consumers, such as delivery distance or temporal differences, such as ride or delivery time;Based on publicly disclosed eligibility criteria to all persons that meet the criteria, such as consumers purchasing in volume, or to all members of a broadly defined group of consumers, such as teachers;Afforded on equal terms to all participants in a loyalty, membership, or rewards program or are offered in response to a consumer complaint, service disruption, request for account cancellation, or similar reason;Offered pursuant to a specified needs-based discount program for reduced pricing related to income or financial need, such as hospital discounted care;Based on a subscription or other continuous agreement that includes a monthly or other recurring price that was not informed by a PWSA; orBased on a refusal to extend credit on specific terms or to enter into a financial transaction based on a consumer's data in a consumer report or data required as part of the application for the financial transaction.     A person has not engaged in individualized wage setting if the person can demonstrate, as described in the act, that the person offers individualized wages based solely on data specific to an individual worker that is directly related to worker seniority or the tasks the worker was required to perform, and the person discloses to the worker before hiring, and to all workers whose wages are set in whole or in part by a PWSA, what data is considered and how the PWSA considers the data.     A person that uses a PWSA shall develop and publish reasonable procedures to ensure the accuracy of all data considered by the PWSA, for workers to request and receive information about what data is collected, and to correct or challenge data considered by a PWSA.     A violation of the prohibition against individualized price or wage setting is a deceptive trade practice under the 'Colorado Consumer Protection Act' and is subject to the enforcement provisions and remedies provided in that act.(Note: This summary applies to this bill as enacted.)

On and after January 1, 2027, when determining coverage for health-care services, the act requires entities that use an artificial intelligence system (AI system) for the purpose of conducting utilization review of health-care services, including health insurance companies (carriers), pharmacy benefit managers, private utilization review organizations, behavioral health administrative services organizations, and managed care entities (entities), ensure that the AI system complies with certain requirements specified in the act. Specifically, an entity shall ensure that the AI system:Makes determinations based on medical or clinical history, the patient's individual clinical circumstances, and other relevant clinical factors specified in the act, with denial of coverage reviewed by a licensed clinician or physician or other competent regulated professional who is competent to evaluate the specific clinical issues and review the health benefit plan's terms of coverage (competent regulated professional);Does not base its determination solely on group data without reference to the individual's data;Is not used in any way that discriminates against individuals in violation of other state or federal laws and is fairly and equitably applied, including in accordance with regulations and guidance issued by the federal department of health and human services; andIs periodically reviewed to ensure the AI systems outcomes are accurate and reliable and that an individual's health data is not used beyond its intended or stated purpose.     Entities that use AI systems shall disclose to the division of insurance, the department of human services, or the department of health care policy and financing, as applicable, the utilization review functions for which the AI system will be used and the points in the utilization review process when it will be used, the process for human oversight of adverse coverage determinations, and the process for maintaining audit information to ensure that the use of the AI system complies with the requirements in the act.     The AI system may be used to assist in utilization review, including expedited approvals. A carrier's denial of coverage for a service based in whole or in part on medical necessity shall not be issued solely on the output of an AI system without human review by a licensed clinician or physician or other competent regulated professional.     Further, the act prohibits a carrier and a payer of services under the 'Colorado Medical Assistance Act' and the 'Children's Basic Health Plan Act' from paying for psychotherapy services that are provided directly to a client and that are conducted by an AI system.(Note: This summary applies to this bill as enacted.)

The bill criminalizes the use of an individual's fingerprint, voiceprint, retina, iris, or facial map (personally identifying feature) in an advertisement, deepfake, image, video, voice recording, or other digital depiction, without the individual's permission (unlawful use of a personally identifying feature). Unlawful use of a personally identifying feature is a class 5 felony, and unlawful use of a personally identifying feature with the intent to harm the individual or another individual is a class 4 felony.An individual harmed by an unlawful use of a personally identifying feature may bring a civil cause of action and collect damages and reasonable attorney fees.(Note: This summary applies to this bill as introduced.)