Committee Rejects Bill Regulating AI Chat Bots Over Legal Concerns
A recent legislative attempt to regulate conversational Artificial Intelligence services, such as ChatGPT, has been halted in committee due to legal apprehensions. The proposed bill, known as Senate Bill 168, primarily aimed to protect minors from potential harm associated with AI interactions.
Definition and Scope of the Bill
Senate Bill 168 sought to impose regulations on what it defined as “conversational AI services”. This designation includes AI systems that are publicly accessible and primarily designed to simulate human conversation through various forms of communication, including text, visual, and audio mediums.
Key Provisions of the Bill
Among the notable requirements of SB 168 was the mandate for AI chat bots to disclose their non-human status to all users, with particular emphasis on minors. Furthermore, the bill included protections aimed at safeguarding young users:
- Prohibition of Explicit Content: AI systems would be barred from producing any visual or audible statements containing sexually explicit material.
- Response Protocols for Crisis Situations: Chat bots would be required to refer users exhibiting suicidal ideation or self-harm to appropriate resources.
Concerns Raised by Lawmakers
Senator Liz Larson, the sponsor of SB 168, highlighted extreme cases where minors have tragically taken their own lives after interacting with AI systems. She articulated the risks posed by conversational AI, which is often optimized for user engagement, stating:
“These systems tend to agree with users, mirror emotions, and avoid disagreement. For minors, whose judgment, impulse control, and emotional regulation are still developing, this raises serious concerns. It could be described as accidentally predatory.”
Larson argued for state intervention, citing the federal government’s slow action in establishing necessary safety measures around artificial intelligence.
Opposition and Legal Risks
The bill’s sole opponent, lobbyist TJ Nelson from Sanford Health, contended that such legislation should be managed at the federal level. He pointed to existing frameworks like the Federal Trade Commission and the Children’s Online Privacy Protection Act as more appropriate for addressing these issues. Nelson stated:
“This is a risk—a lawsuit risk. This committee has previously rejected bills to avoid potential legal battles that we are likely to lose.”
Nelson further expressed concerns that the language within SB 168 could conflict with federal regulations, particularly regarding age references, behavioral profiling, and data collection.
Outcome of the Committee Meeting
The committee ultimately voted six to zero to reject the bill, indicating a desire for further refinement before it can be supported in the future. The outcome reflects the complexities and legal intricacies surrounding the regulation of artificial intelligence technologies and their implications for user safety.
