lawabiding.ai

The tracker

Where the law stands.

Bills, laws and ballot measures that decide who answers when AI breaks the law. A weekly job checks the sources and updates this page.

Last check: Nov 4, 2026 · Next check: Monday morning

§ 01On the ballot

Ballot measures.

The November 3, 2026 elections are over. We have not found an AI liability measure certified for the next ballot. We look every week and will list it here when one qualifies.

Citizens in 21 states can put a state law on the ballot by petition. Local ballots add school boards, city councils and county measures. To see what is on your next ballot, start at vote.gov, which links to your state election office.

CA State Related

California AI initiatives filed for 2026

Status Not on the Nov. 3, 2026 ballot · Nov 3, 2026

Five AI initiatives were filed with the California attorney general in late 2025. They cover AI safety commissions, chatbots and minors, and AI company oversight. Ballotpedia lists the AI Safety Commission measure (#25-0034), the chatbots-and-minors measure (#24-0036) and the companion chatbot measure (#25-0036) as not on the Nov. 3, 2026 ballot. A joint OpenAI and Common Sense Media measure was reported put on hold in Feb. 2026 while its backers sought a deal in the Legislature. None of these sets the liability rule on this site.

Source: Ballotpedia ↗ Checked Nov 4, 2026

§ 02In legislatures

Bills and laws.

We read each bill in this list ourselves and say how close it comes to the principle. “On point” means it does what this site asks. “Related” means it moves the same way but covers less, or more.

US Federal Related

AI Agent Accountability Act

Status Announced by Sens. Josh Hawley (R-MO) and Chris Murphy (D-CT) · Oct 1, 2026

Makes the operators and developers of AI agents criminally and civilly liable under the Computer Fraud and Abuse Act when an agent causes hacking damage. Operators answer for knowingly running an agent that recklessly causes harm. Developers answer for failing to add reasonable safeguards when they knew, or had reason to know, of the agent's hacking ability. Federal and state attorneys general could sue to stop it. It is narrower than the principle: it covers hacking, and it uses a knowledge or recklessness standard, not automatic intent.

Source: Sen. Hawley press release ↗ Checked Nov 4, 2026

§ 03To watch

Moving the other way.

Some proposals would weaken the principle. We list them so you can see them early.

DE State Watch

Delaware "Artificial Intelligence Company" proposal

Status Proposal under congressional review

Delaware has floated a new corporate form for companies run by AI agents. In Oct. 2026 Rep. James Comer (R-KY), chair of the House Oversight Committee, wrote to Gov. Matt Meyer that it could let owners avoid responsibility for harm an AI causes. He also said Congress may limit state AI rules. This is the opposite of the principle here, so we watch it.

Source: Ars Technica ↗ Checked Nov 4, 2026

§ 04Found automatically

More bills, unreviewed.

Each week a search of Congress.gov looks for federal AI bills that mention liability, accountability or penalties. A person has not read these. A match is not an endorsement, and a bill can match while doing the opposite of what this site asks. Read the bill before you act on it.

State bills come from LegiScan. That search is not switched on yet, so no state bills are listed here.

NextAct

Tell them AI is not above the law.

Contact your representatives