As artificial intelligence expands its reach into more aspects of people’s lives, Colorado lawmakers want to ensure the state’s policies around online harassment and child pornography are keeping pace.
This story was produced as part of the Colorado Capitol News Alliance. It first appeared at cpr.org.
“In a modern world where our identities and our selves exist far beyond our physical bodies and are vulnerable to attack on media, on social media, in text messages… we have to adapt to protect the people in these spaces,” Jessica Dotter, sexual assault resource prosecutor with Colorado’s District Attorneys Council, told the Senate Judiciary Committee earlier this week.
She was there to testify in favor of Senate Bill 288. The legislation, sponsored by Majority Leader Robert Rodriguez, a Denver Democrat, would expand the existing laws around posting intimate images to include material created by AI.
The issue exploded into the public conversation last winter, when deepfakes of megastar Taylor Swift emerged online. But the creation of false sexual pictures and videos has torn apart the lives of much less famous people, including a city commissioner in Florida and a teacher in Texas and high school students nationwide.
“Let’s be clear: These images are not harmless. They’re not virtual, they’re violations,” testified Will Braunstein, the head of the Denver Children’s Advocacy Center.
According to backers, Colorado is late to tackle this issue. While the issue of obscene deep fakes is relatively new, states have been moving quickly to criminalize the practice, when it’s used to harm someone. Thirty-eight states have made it illegal to use AI to create child pornography, something Colorado’s bill would also do.
That provision is raising concerns with both the ACLU of Colorado and criminal defense attorneys, who question the ethics of prosecuting people for possessing or distributing material entirely created by a computer.
“When we talk about someone under the age of 18 and it’s not tied to a real person, how do you put an age on that?” James Karbach with the state’s Public Defenders Office asked lawmakers. “How do you draw lines around that? And if it’s to be criminalized, should you criminalize it the same as something that harms and depicts an actual human?”
The Daily Sun-Up podcast | More episodes
Bill supporters countered that AI is trained on real images with real victims. And they note that as the technology improves, it’s becoming increasingly difficult to separate deepfakes from real images of abuse, making it more important — in their arguments — to treat both the same under the law.
The debate over abusive deep fakes comes as Colorado lawmakers are generally trying to get their arms around emerging technology and its potential impacts on people’s lives. Rodriguez is working behind the scenes on even more sweeping legislation: a first-of-its-kind effort to put guardrails around how companies use AI to make decisions with consequences for people’s lives.
That bill has yet to be unveiled.
In the meantime, Rodriguez’s exploitative content bill passed its first committee unanimously and is awaiting a vote of the full Senate.