Regulating artificial intelligence has reached a critical stage in California, as lawmakers have proposed a bill regarding children’s safety restrictions on AI chatbots called the Leading Ethical AI Development (LEAD) for Kids Act, or also known as Assembly Bill 1064. According to the California Department of Justice, the bill was presented by Assemblymember Rebecca Bauer-Kahan because more and more people are gaining concern for the influence artificial intelligence has on young people, especially when the companies running these softwares are aiming to spread their technology globally. The California Governor has until October 12th to make his decision about the bill.
The bill’s goal is to protect children from manipulative advice, privacy risks, and potentially harmful interactions with these AI systems, which are designed to imitate human relationships and conversations. If the bill was passed, it would greatly affect major tech companies such as OpenAI and Google, banning them from making certain chatbots available to minors unless they have specific regulations to prevent such risks. These may include making sure that the AI is incapable of encouraging self harm and providing unsupervised mental health advice. The law would overall establish accountability for AI companies to make sure that the developers prioritize the well being of younger individuals instead of simply pure data collection.
The efforts to pass this bill has created a great deal of debate between supporters and opposition. Those who advocate for the bill, such as parents and victims, argue that the bill aims to protect children from any potential harm from AI interactions, while others like the major tech companies believe that it might restrict children’s access to helpful tools, especially in the modern era with the increase in technology. Privacy concerns also rise because of how much data AI companies have already been collecting. “On the plus side, any regulations online are helpful,” Ms. Hight, an AP Government and Ethnic Studies teacher at Wilcox High School says. “There’s currently minimal regulations preventing businesses and AI programs from specifically targeting children, and many online platforms are designed to target children.” Those who advocate for this bill believe that the law helps to define the boundaries for acceptable behavior from these companies when younger users are involved.
On the other hand, critics raise questions about how this law will actually be enforced and the potential breaches of privacy. Ms Hight notes, “[I] can’t say how effective this sort of bill would be. Legislation isn’t effective unless it is enforced, and there doesn’t seem like a current plan to enforce the decision.” Additionally, people worry that the bill infringes on the freedoms that the AI companies have and limits children from helpful tools, especially in this “new era” of increased artificial intelligence. Furthermore, organizations like SIIA argue that the bill’s content restriction could infringe on people’s First Amendment rights. “The law often sides with safety over personal freedoms,” Ms. Hight notes. “This is more about regulating how businesses collect individual information… if the regulations aren’t clearly defined and enforceable, then nothing will actually be accomplished which could lead to pushback against future regulations.”
Surprisingly, the bill connects to a broader national struggle to define AI’s legal limitations. If the bill is passed and is enforceable, it could change how the public actually views AI. Ms. Hight claims that “it could put AI in a box where it’s recognized as a business tool, similar to algorithms in social media.” She believes that this clarification is important because “a lot of students [she] speaks with think that AI exists separately and is something that actually speaks back, instead of being a computer program that is good at pattern recognition owned by a business.” By providing legal limitations to AI, it highlights how the bill forces people to understand these types of technology are services that companies profit off of and not an individual entity.
While the bill might not seem as important to those who live outside of California, its potential existence is a very powerful signal for the future. “I do think legally it could be better placing AI into an already existing model that people can use, which might increase understanding,” Ms. Hight claims. Furthermore, she says that “if the legislation is harsh, that could also lead to people looking into the dangers around AI more, which could lead to better understanding.” Ultimately, this California bill, by addressing the protection of younger individuals online, shows how the country is attempting to take steps to regulate AI, but whether or not these steps will be effective and efficient is the uprising debate.
