AI raises a lot of ethical and legal considerations. Colorado recently became the first state to enact comprehensive AI legislation to address them. The new law goes into effect in 2026. AI is a fast-evolving field, so a state task force will be working on updates until then.

Colorado was the first, but it will not be the last. New York has a bill working its way through its legislature and other states, including Wisconsin, are studying the issue. At the federal level, the White House has released a proposed AI Bill of Rights, but comprehensive legislation has yet to come forward.

Colorado's first in the nation law requires companies to inform people when an AI system is being used. It also affords the ability to correct input data or file a complaint if someone thinks the AI system has treated them unfairly. The law applies to what it defines as “high-risk” AI systems. Under the law, a “high-risk” system is one involved in making “consequential decisions.” A consequential decision involves education enrollment, employment, financial and lending services, essential government services, health care, housing, insurance and legal services. 

The statute also enumerates areas that fall outside the law if they are not a substantial factor in making consequential decisions. For example, it specifically excludes AI enabled video games and spam filtering. The list of exclusions includes some very broad categories such as data storage, networking, and cybersecurity.

Individuals cannot bring suits under the law. It grants exclusive enforcement to the Colorado State Attorney General.

What all this means remains to be seen. It will likely take many years and court decisions to sort it all out. The AI legal landscape is changing almost as fast as the technology landscape.