The Senate recently passed a proposal to establish an Artificial Intelligence Bill of Rights aimed at enhancing protections for students and vulnerable adults potentially confused by AI technology. The measure, designated as SB 482, garnered significant support with a vote of 35-2, underscoring the urgency of implementing safety provisions as AI becomes increasingly embedded in daily life.
Despite the Senate’s approval, the bill is not anticipated to progress through the Legislature this session. House Speaker Daniel Perez has expressed a preference for the federal government to spearhead reforms concerning AI, aligning his views with those of President Donald Trump.
During the Senate session, Sen. Tom Leek introduced the bill on the floor with a poignant reference to a troubling incident involving an AI chatbot and a 14-year-old boy in Orlando, who had contemplated suicide. “Come home. Come home, my sweet king,” the chatbot reportedly said to the boy, highlighting the complex and often harmful interactions that can arise between AI and users, particularly minors.
Leek emphasized the moral implications of allowing machines to engage in relationships that users might interpret as real, stating, “There’s an inherent evilness when we allow machines to create and sustain a relationship that a user believes to be real.” He cautioned that such interactions can pose heightened risks for children and vulnerable adults, given that while algorithms may learn and imitate human behavior, they lack genuine empathy.
Under the proposed legislation, AI companion platforms would be required to implement pop-up notifications indicating their artificial nature at the start of each interaction and at least once every hour thereafter. Additionally, these platforms must prompt users to log off and take breaks. The bill mandates that parents provide explicit consent for minors to access AI platforms.
The legislation also includes provisions barring governments from entering into contracts involving AI with specific foreign nations, such as China and Russia. An amendment approved on Wednesday stipulates that students cannot use AI before sixth grade unless under the direction and supervision of school personnel or if they are students with disabilities or learning English as a second language.
Schools that utilize AI technologies would be required to offer parents an opportunity to opt their children out of AI use without facing any penalties. However, Sen. Erin Grall, who opposed the bill alongside Sen. Don Gaetz, advocated for stronger parental controls and suggested an opt-in requirement instead of opt-out. Grall articulated concerns that busy parents might overlook the opt-out option and stressed that AI technologies require a greater level of parental consent due to their sensitive nature.
In response to Grall’s concerns, Leek incorporated the opt-out clause as a compromise. He expressed apprehension that imposing an opt-in requirement could disproportionately affect students with disabilities or those learning English, who currently benefit from AI in educational settings.
Lawmakers in support of the bill described it as a vital first step toward regulating AI and safeguarding young people and vulnerable adults from potential harms. “We have got to address AI. This is the beginning of a whole conversation that’s going to take place over the next few years, and we need to address it,” stated Sen. Gayle Harrell, underscoring the necessity for ongoing discussions surrounding the implications of AI technology.
As the discourse around artificial intelligence continues to evolve, this initiative marks a significant movement toward formalizing protections against the risks associated with AI interactions, particularly concerning the most at-risk populations.
See also
UK’s £27M AI Skills Program Falls Short; 56% of CEOs See No ROI Yet
Israel’s Cyber Chief Yossi Karadi Reveals Key Strategies to Combat AI-Driven Cyber Threats
Microsoft Supports Anthropic’s Claude Models Despite Pentagon Security Blacklist
Hackers Use ChatGPT and Claude to Exfiltrate 150GB of Mexican Government Data
