Lawmakers Propose Penalties for Digitally Manipulated Sexual Content

Hawaii legislators are advancing measures to protect residents from AI-manipulated imagery and automated decision-making systems amid growing concerns about personal privacy and digital exploitation.

For Ka’iulanihalimakio Abeita, an 18-year-old model attending the University of Hawaii-Hilo, the threat of manipulated images is a constant worry. “For someone to just create a deepfake that could take off my bikini,” Abeita said, “things like that cross my mind all the time.”

Senate Bill 1156 aims to address these concerns by classifying the creation and distribution of “sexually explicit” deepfakes as a misdemeanor. The legislation defines deepfakes as altered media that would appear “to a reasonable observer” to be an authentic record of a person’s actions or speech. The bill contains exemptions for news content, satire, parody, and material that clearly discloses it is not depicting a real event.

In parallel, lawmakers are considering Senate Bill 2967, which focuses on regulating AI systems in consumer interactions and “high-impact decisions.” The bill acknowledges that artificial intelligence has become “integral to commerce, education, health care, finance, transportation and government operations,” and would require businesses to disclose when consumers are interacting with AI systems.

Rae Sunshine, a 25-year-old creative strategist, believes the legislation addressing sexually explicit deepfakes is long overdue. Even before AI’s rise, Sunshine says better enforcement against image manipulation was “something that’s been necessary.”

According to Sunshine, deepfake imagery shifts focus away from a woman’s character or work and instead “weaponizes their body and sexuality.” She has personally experienced the harm of manipulated media when audio from her Instagram videos was edited to make it appear she said something controversial.

“It can have a very negative effect on someone, even with less than 10 seconds of material,” Sunshine explained. “What happens online, it reverberates like a butterfly effect and it affects you in real life.”

The Department of the Attorney General has expressed support for the intent behind SB 1156, noting it aligns with their goal of protecting individuals. However, in written testimony, the office indicated that existing law already criminalizes the creation or distribution of “deepfake images or videos intended to harm a person.”

For models in Hawaii’s limited market, the stakes are particularly high. Kyani Bateman, a 20-year-old model and social media coordinator for Waikiki Beach Services, emphasizes that consent is fundamental in the modeling industry.

“Unless it’s an actual written ‘yes’ that both of you guys have agreed on, it should not be happening,” Bateman said. She warned that manipulated images can permanently damage reputations even after they’re proven false, noting that “your face is your brand, your image is your career.”

Beyond the modeling industry, young people across Hawaii are concerned about the spread of manipulated media. Arisa Yazaki, a 19-year-old student and dance teacher, points out that students affected by deepfakes often “don’t know the resources they have to fix these problems.”

The potential consequences extend beyond personal embarrassment to professional opportunities. Abeita explained that deepfakes could affect her ability to secure brand endorsements and deals, emphasizing that before SB 1156 was introduced, there were few consequences for those who created sexually explicit images of real people.

“They could just, essentially, get away with it,” Abeita said.

As AI technologies become more sophisticated and accessible, the implications reach far beyond celebrities and public figures. “This could happen to somebody’s kids… somebody’s mom, somebody’s grandma,” Abeita warned. “This doesn’t just affect big names.”

The legislation represents Hawaii’s effort to address growing concerns about AI ethics and digital security in an era when the line between authentic and manipulated content continues to blur.