Artificial intelligence chatbots operating in Colorado would be required to adhere to a series of regulations aimed at protecting kids and preventing suicide under a bipartisan bill introduced in the legislature last week.
State Rep. Sean Camacho, a Denver Democrat, said he brought House Bill 1263 in response to outcry from his constituents, some of whom reported that their children were “sexually groomed” by chatbots before going on to harm themselves.
“Other than the Taxpayer’s Bill of Rights, this is the No. 1 issue people wanted to talk about in the offsession,” he said in an interview Tuesday, referencing the interim since lawmakers last gathered at the Capitol.
The measure is the legislature’s latest attempt to address artificial intelligence as the technology becomes increasingly prevalent. In addition to the new chatbot bill, lawmakers are expected to revisit their 2024 AI disclosure law — the first of its kind in the nation — which is scheduled to go into effect in June after two previous failed attempts to tweak it.
The challenge for state lawmakers has been trying to figure out how to regulate AI in Colorado without stifling innovation. The legislature has also been trying to balance how those state regulations would function when AI companies operate nationally.
Starting in 2027, House Bill 1263 would require companies operating AI chatbots in Colorado to “clearly and conspicuously” notify children using their platforms that they are interacting with AI and not a real person. It would also prohibit the platforms from offering rewards to increase engagement with children.
The measure would also require the platforms to “institute reasonable measures” to prevent their AI chatbot from producing sexually explicit content — either images or text — when asked to by a child user. Additionally, the legislation would require tech companies to take reasonable measures to prevent a chatbot, when interacting with a child, from “generating a statement that stimulates emotional dependence,” such as romantic role-playing.
The bill also seeks to require companies to allow for parental controls if their AI chatbots are accessible to kids.
Finally, the legislation would require that AI chatbots provide suicide-prevention resources to users who express thoughts of suicide or self-harm. Platforms would be required under the bill to file reports with the attorney general’s office on how often a chatbot flags suicidal or self-harm behavior.
Violations of the bill would be treated as a Colorado Consumer Protection Act infraction, punishable by a fine of up to $1,000 per occurrence.
Camacho is leading the push for the bill in the House alongside state Rep. Javier Mabrey, a fellow Denver Democrat. The bill’s main sponsors in the Senate are Sens. Iman Jodeh, D-Aurora, and John Carson, R-Highlands Ranch.
Camacho said the sponsors worked with tech companies to try to come up with a bill they could reasonably implement.
“I think they are open to it,” he said, adding that he thinks the measure is narrowly tailored enough that tech companies will be able to adhere to rules in Colorado without affecting their operations across the country.
A handful of states have already enacted laws around AI chatbots, including California, Maine and New York. Camacho said House Bill 1263 incorporated parts of those other states’ measures.
