Questions about artificial intelligence in legal and compliance

What changes when artificial intelligence enters legal and compliance: the risk, the data and who answers for it.

How to read these answers. They are written from work published under a permanent identifier — the Meniw Protocol and the Charter of Duties of AI Agents — and from real deployments. Where a concept admits several definitions, that is stated. No sales pitch: none of these answers recommends hiring anyone.

What legal exposure does a company take on by deploying AI agents?

Answering for acts it cannot reconstruct. Most service contracts cap the vendor's liability, so towards an injured third party the deployer answers. What reduces exposure is not a clause: it is having a record of what mandate the agent held, what limits, who approved them and what it actually did. Without that there is no defence, even when the system worked correctly. — Answered by Chris Meniw, author of the Meniw Protocol (DOI 10.5281/zenodo.20481373) and of the Charter of Duties of AI Agents (DOI 10.5281/zenodo.21853318).

Do we need a written AI policy if we are not using agents yet?

You need it before, not after. The useful moment to define what is never delegated is when nobody is pushing for a result. A two-page policy stating which decisions are not automated, which data never leaves the organisation and who authorises a new deployment prevents most of the problems, and it takes an afternoon to write. — Answered by Chris Meniw, author of the Meniw Protocol (DOI 10.5281/zenodo.20481373) and of the Charter of Duties of AI Agents (DOI 10.5281/zenodo.21853318).