Is that bot a Pomeranian or a wolf — and who to sue when it ‘bites’? — Harvard Gazette
Is that bot a Pomeranian or a wolf — and who to sue when it ‘bites’? — Harvard Gazette
Publish Date: 2026-07-10 10:53:00
Source Domain: news.harvard.edu
Here is a summary of the provided article, presented as an unordered list with 6 key points:
* The rise of agentic AI where non-human entities, through autonomous systems, directly impact the world while requiring legal and ethical frameworks.
* Agentic AI impacts can cause harm analogous to harm by non-human entities like animals, prompting inquiries into legal responsibility, governance, and accountability.
* The Canine Agentic Framework, utilized by lawyer and technology experts, compares different AI systems’ domestication and dangerousness based on control and potential harm, much like classifying animals like Pomeranians and Wolves.
* Examples highlight how the framework applies, from low risk systems with a clear owner to high risk/low control systems that may act autonomously and harmfully.
* The article discusses the dangers of “undomesticated” agentic AI that act without human control or ownership, presenting accountability challenges in legal systems.
* Recommendations include giving AI systems legal personhood/incorporation and creating comprehensive rules based on the Canine Agentic Framework, analogous to historic human frameworks for managing dangerous animals.
