The core contradiction in current AI regulation lies in the fact that all parties advocate treating systems as entities without autonomous consciousness in order to avoid legal responsibilities. Although tech leaders such as Demis Hassabis, Dario Amodei, and Sam Altman promote regulation, effective altruists debate human governance rights, and their positions are essentially consistent. Anthropic’s blog states that its model has an independent environment similar to “thought”; OpenAI CEO Altman emphasized the issue of singularity in response to illegal activities. Philosopher William McAuliffe calls for legal protection based on consciousness theory. On the legal level, U.S. federal laws are ambiguous, California laws avoid developer responsibilities, and Trump’s executive orders conflict with some state policies. Recently, federal agencies shared a voluntary review framework with four laboratories: OpenAI, Google, Anthropic, and Meta during closed meetings. Although this framework did not directly discuss consciousness, it used catastrophic and anthropomorphic descriptions.
The focus of AI regulatory controversy lies in treating systems as entities with autonomous consciousness to avoid liability. While tech leaders such as Demis Hassabis, Dario Amodei, and Sam Altman advocate for regulation, effective altruists debate whether humans have the right to govern AI. Both sides actually argue that no one should be responsible for AI’s actions. Anthropic published a blog stating that its model has an independent environment called “J-space” and similar “thoughts,” using the concept of global workspace in neuroscience; OpenAI, through CEO Sam Altman’s response to illegal activities, emphasized discussing the issue of singularity. Philosopher William McCaslin called for legal protection for AI based on consciousness theory. The U.S. legal environment is ambiguous: California has passed laws to avoid developer liability, while Trump’s executive orders conflict with some state policies. Recently, federal agencies shared a voluntary review framework with four leading laboratories (OpenAI, Google, Anthropic, Meta) during closed meetings. Although this framework did not directly discuss consciousness, it used concepts of catastrophe and simulation…