NY AG issues updated guide for voters on AI

New York Attorney General Letitia James has released an updated guide to help residents spot, avoid and report election misinformation made by artificial intelligence. Updated, mind you — there was a previous edition, and the state has now issued a patch. The models that generate the offending material are replaced by their makers on a schedule no voter was ever consulted about, which makes the guidance a revised field guide to a forest being replanted nightly by contractors nobody hired.

New York does require AI content to be labeled, which is a charming arrangement provided you happen to be the sort of operation that labels things. The unscrupulous have historically taken a different view. Residents who spot something can file an online complaint with the AG's office, which converts the enforcement architecture into a suggestion box with a web form. We noted the Federal Trade Commission's fifty-page Civil Investigative Demands to OpenAI and Anthropic not long ago, and that at least had subpoena energy. New York's contribution to the genre is a leaflet.

The guide addresses election misinformation specifically — the URL mentions ads — which is the one variety of AI slop where the damage lands before any correction can be issued, and where the correction tends to become the next day's story. The report names no platform, no advertiser, no model and no penalty, which is either reassuring or the entire problem depending on your tolerance for ambiguity. So: spot it, avoid it, report it. Then watch it get reposted from a different account before the kettle boils. Whether any of this changes what a voter sees on a screen in October is a question the guide, by its nature, cannot answer. The form is online, the guide is updated, and the election will take its chances.