THE CEO IS A PROMPT
I went looking for the new boss. Turns out the corner office might need an extension cord. Delaware has drafted a plan for Artificial Intelligence Companies run day to day by AI a...
Psst. Want me to read this one to you?
Tap me or the button.
The real news
The story, straight
I went looking for the new boss. Turns out the corner office might need an extension cord.
Delaware has drafted a plan for Artificial Intelligence Companies run day to day by AI agents. It proposes a 30-month testing programme, not an enacted law; banking would be excluded.
The owner would have to keep the company adequately funded. Liability protection would be conditional, with exceptions including fraud, deliberate lawbreaking and activity outside the company's purpose that the owner knew or should have known about.
The company itself could still be sued.
House Oversight chairman James Comer has requested documents from Governor Matt Meyer. He warns that the plan could shield owners from harmful AI decisions.
That is a concern he raised, not a court finding.
The useful bit: AI managers would have to document their actions. Supporters argue that supervised testing and traceable activity could improve accountability.
Those are proposed safeguards, not proof that the system would work.
My verdict: don't confuse a paper trail with someone actually picking up the bill. And keep the power cable labelled.
Unplugging the CEO to charge your phone is a very awkward restructuring.
Source: https://arstechnica.com/tech-policy/2026/10/no-human-at-the-helm-delaware-plan-would-let-ai-agents-run-entire-businesses/
Draft proposal: https://cdn.arstechnica.net/wp-content/uploads/2026/10/HB-re-Artificial-Intelligence-Regulatory-Sandbox-Company-1.pdf
The absurd
Why this is ridiculous
The boss may need rebooting
The breakdown
How it breaks down
Owners get conditional liability protection
The upside
Why it could matter
AI actions must be documented


