CoolFace
Datasetpublic

CyberMax-tools/policywren-us-ai-policy-tracker

Policywren: US AI Policy Tracker Need it fresh, filtered or via API? This free file is a snapshot (US AI policy documents up to the last refresh), last updated 2026-09-25. Swellmeter on Apify ($0.003 per keyword analyzed): checks Google Trends direction for your own keywords with a plain rising/falling verdict, daily if scheduled. Get an email when this dataset updates: free, double opt-in, unsubscribe any time. Every document in the US Federal Register since… See the full description on the dataset page: https://huggingface.co/datasets/CyberMax-tools/policywren-us-ai-policy-tracker.

sourceHugging Facecc0-1.0updated 10h agoView on Hugging Face
0likes65downloads
Dataset Card

Policywren: US AI Policy Tracker

<!-- cm:banner --> Policywren: Every US federal AI rule and comment deadline <!-- /cm:banner -->

<!-- cybermax-bridge:start -->

Need it fresh, filtered or via API? This free file is a snapshot (US AI policy documents up to the last refresh), last updated 2026-09-25. - [Swellmeter on Apify](https://data.cybermax-tools.workers.dev/buy/swellmeter?s=hf-policywren-us-ai-policy-tracker) ($0.003 per keyword analyzed): checks Google Trends direction for your own keywords with a plain rising/falling verdict, daily if scheduled. - [Get an email when this dataset updates](https://data.cybermax-tools.workers.dev/notify?ds=policywren-us-ai-policy-tracker&s=hf-policywren-us-ai-policy-tracker): free, double opt-in, unsubscribe any time.

<!-- cybermax-bridge:end -->

Every document in the US Federal Register since 2017-01-01 whose text mentions artificial intelligence: executive orders and other presidential documents, final rules, proposed rules, notices, requests for information and their comment periods, with dates, agencies, dockets and links. Rebuilt daily from the official Federal Register API. Neutral and factual; not legal advice.

This build (2026-09-24 17:23 UTC): 1,401 documents (935 notices, 209 rules, 189 proposed rules, 68 presidential documents); 253 are mainly about AI (AI in the title, action or summary) and the rest mention it in the body; 11 AI executive orders and presidential documents; 18 comment periods still open; latest publication 2026-09-24.

configrowswhat
documents (default)1,401one row per Federal Register document, newest first
open_comment_periods18documents whose comment period closes today or later, soonest first
public_inspection0AI-related documents filed for publication but not yet published (today's public inspection list)
weekly_counts823documents per week, by type and ai_focus

documents columns: document_number, publication_date, type (Rule, Proposed Rule, Notice, Presidential Document), subtype (e.g. Executive Order), executive_order_number, title, action, abstract, agencies, agency_slugs, ai_focus (primary = AI in title/action/summary, mention = elsewhere in the text), matched_terms, significant (EO 12866 significance, when given), comments_close_on, effective_on, signing_date, docket_ids, regulation_id_numbers (RIN), topics, citation, excerpt (the passage that matched), html_url, pdf_url, comment_url. Also as data/documents.csv.

Search query: "artificial intelligence" | "machine learning" | "generative AI" | "large language model" | "super intelligence" | superintelligence | "frontier model". A document that only mentions AI once is still included, so filter on ai_focus == "primary" for documents about AI. Later orders can revoke or amend earlier ones; the Federal Register page of each order shows its status.

<!-- cm:gallery --> Policywren: product image 1 Policywren: product image 2 Policywren: product image 3 <!-- /cm:gallery -->

Use it

python
from datasets import load_dataset
docs = load_dataset("CyberMax-tools/policywren-us-ai-policy-tracker", split="train").to_pandas()
eos = docs[(docs.type == "Presidential Document") & (docs.ai_focus == "primary")]
open_now = load_dataset("CyberMax-tools/policywren-us-ai-policy-tracker", "open_comment_periods", split="train").to_pandas()

Source and license

Federal Register content is a US government work in the public domain; this compilation is released under CC0 by CyberMax. Built with the public Federal Register API (no key). Corrections: open a discussion on this dataset.

<!-- cybermax-notebook:start -->

Quickstart notebook

Load this dataset with pandas and see real outputs in the quickstart notebook. <!-- cybermax-notebook:end -->

<!-- cybermax-web:start -->

Browse it on the web

Free, no sign-up: a feed of new AI policy documents: RSS · JSON. AI agents can query it through the remote MCP server https://data.cybermax-tools.workers.dev/mcp (no key). <!-- cybermax-web:end -->