Artificial intelligence, long on the mind of tech entrepreneurs, finally made a splash with the general public with the release of numerous consumer-facing generative AI technologies. AI images, text and videos are now being produced at a breakneck pace and are appearing in all facets of public life.
As it stands, there are few guardrails and laws surrounding the use of AI. Colorado lawmakers are looking to take some steps in the direction of regulation with three bills working through the Colorado legislature. However, the bills are in a race against the clock as this year’s legislative session approaches its end.
A Guard Against Deepfakes in Elections
The first bill, and the closest to becoming law, is House Bill 24-1147, titled Candidate Election Deepfake Disclosures. The measure “creates a statutory scheme to regulate the use of deepfakes produced using generative artificial intelligence in communications about candidates for elective office.”
The bill doesn’t ban the use of generative AI outright, rather it adds a disclosure requirement to their use in certain cases and creates a civil private right to action when that requirement isn’t met.
The bill also carves out several protections, including protections for news media, websites and media content that constitutes satire or parody.
Rep. Brianna Titone, one of the bill’s sponsors, told the committee the bill’s goal was to make sure any deepfake media images, video clips or audio clips have some kind of disclosure attached to them.
“So that way anyone who may be consuming this information that is put out in the air has the ability to tell what has been put out that is not a real representation of what the truth is,” said Titone. “It’s not an easy task.”
Rep. Junie Joseph, the bill’s other prime sponsor in the House, told the committee the bill was supported by the Colorado Secretary of State. Joseph also noted the sponsors had been in conversations with various organizations and stakeholders on the bill, which led to seven amendments in the initial committee hearing.
Christopher Beall, Colorado’s deputy Secretary of State, testified at the hearing. He told the committee the bill, as introduced, would have complaints of this nature go through the existing campaign finance complaint process.
Aly Belknap, executive director at Colorado Common Cause, testified in support of the bill and said the availability of generative AI has made it easier than ever before to spread false information and propaganda.
“In this increasingly complex information ecosystem, it’s critical that citizens have the tools to determine whether an image, video or audio representation made by a candidate campaign is authentic and truthful,” said Belknap.
The President and CEO of the Colorado Broadcasters Association, Justin Sasso, told the committee the association was in an amend position on the bill as introduced.
“The CBA and its members do not feel it goes far enough to exempt broadcasters from false claims of altered content or exposure to frivolous lawsuits that are often made by those simply attempting to stall an opponent’s campaign,” said Sasso.
In total, the bill was amended 12 times from its introduction to its final passage. It passed its final hurdle in the General Assembly on April 30 and now awaits the signature of the governor.
Protection for Consumers in Interactions with AI
The second artificial intelligence bill, Senate Bill 24-205, has been moving fast through the legislature since its initial Senate Judiciary Committee hearing on April 24. It was recently passed with two amendments by the House Committee on State, Civic, Military, & Veteran Affairs, but it joins a long list of bills awaiting a second or third reading by the House.
The bill largely focuses on protecting consumers from algorithmic discrimination in a “high-risk system.” The bill defines a high-risk AI system as “any artificial intelligence system that, when deployed, makes, or is a substantial factor in making, a consequential decision.”
For algorithmic discrimination, the bill defines it as “any condition in which the use of an artificial intelligence system results in an unlawful differential treatment or impact that disfavors an individual or group of individuals based on the basis of their actual or perceived age, color, disability, ethnicity, genetic information, limited proficiency in the English language, national origin, race, religion, reproductive health, sex, veteran status, or other classification protected under the laws of this state or federal law.”
Senate Majority Leader Robert Rodriguez is the prime sponsor of the bill and told the committee that harm and negative outcomes will continue to mount from AI until action is taken.
“At the base of this bill and policy is just accountability, assessments and disclosures that people need to know when they’re interacting with artificial intelligence,” said Rodriguez.
The bill includes mechanisms for AI developers to comply with the law and methods to remedy cases where an AI technology falls afoul of the bill. In addition, the bill grants the attorney general rule-making authority to implement and enforce the requirements of the bill.
The bill met significant pushback in its initial committee hearing with testimony, largely in amend or oppose positions, lasting around three hours. This pushback came from a variety of groups, including a number of interest groups and smaller developers worried about the risk to them and Colorado’s ability to attract technology investment. The AFL-CIO was concerned that the bill creates a weak foundation to build a regulatory framework for AI while the Denver Metro Chamber of Commerce was troubled by the potential consequences to small businesses and technology companies in the state.
Testifying in support of the bill were a group of young students who were concerned about the rapid development and implementation of AI and the current and future problems associated with the technology.
The bill passed narrowly through its initial committee hearing and has been amended five times since its introduction, including an amendment that would give the bill a longer runway to implementation.
A Taskforce on the Use of AI in Facial Recognition
House Bill 24-1468 was the AI bill introduced latest in the session on April 29. The bill comes from the legislature’s Joint Technology Committee and updates the membership and issues of study for the task force for the consideration of facial recognition services and changes its name to the biometric technology and artificial intelligence policy task force.
Notably, the bill significantly increases the scope of the taskforce. As it stands, the task force focuses on government use of facial recognition technology. If the bill passes, that would expand to include biometric technology and AIe as well. It would also broaden the focus to general use of these technologies, rather than just their use by state and local government agencies.
The measure increases the task force from 15 to 17 members. New additions to the task force would be an expert in generative AI, an advocate for individuals who have historically experienced discrimination by AI and facial recognition technologies and an advocate for youth safety and privacy. The representative from the Colorado Bureau of Investigation will be replaced by one of these new members.
The bill’s initial committee hearing was on May 1 in the House Committee on Business Affairs & Labor. Anaya Robinson, representing the ACLU of Colorado, told the committee his organization was in an opposed position on the bill as originally introduced.
Robinson said the issue for the ACLU was in the portion around schools on the bill, which was struck in the amendment passed by the committee. Robinson said with the strike-through, the ACLU would move to a neutral position on the bill.
The bill has since been amended four more times in its second and third reading in the House. While the bill passed its Senate committee hearing on May 6, it only has two days to pass through its readings on the Senate side to make it to the governor’s desk.