Tag
The article argues for regulating the software harness in agentic AI systems instead of focusing on AI model inference, emphasizing human oversight and accountability to prevent misuse.