Privacy Policy & Terms of Service Change Monitor: Diff avatar

Privacy Policy & Terms of Service Change Monitor: Diff

Pricing

from $6.62 / 1,000 policy checkeds

Go to Apify Store
Privacy Policy & Terms of Service Change Monitor: Diff

Privacy Policy & Terms of Service Change Monitor: Diff

From $11 per 1K policies checked. Find out what changed in a vendor privacy policy or terms of service: added and removed sentences, change percent and the date it last changed, compared with the Wayback Machine or your last run. No API key, no proxy.

Pricing

from $6.62 / 1,000 policy checkeds

Rating

0.0

(0)

Developer

Detroit Web Agent - Open Data

Detroit Web Agent - Open Data

Maintained by Community

Actor stats

0

Bookmarked

2

Total users

1

Monthly active users

8 days ago

Last modified

Features

Limited permissions

Share

Find out what changed in a vendor's privacy policy or terms of service. Give the Actor a list of domains (or direct policy URLs) and it finds each privacy policy and terms page, reads the current text and compares it with an archived Wayback Machine copy from a year ago, or with what it saw on your last run. You get one row per document with the added sentences, the removed sentences, how much of the text changed and the date the current wording first appeared.

Typical jobs: vendor risk and third-party reviews, compliance and legal watch lists, checking which SaaS tools quietly changed how they use your data, and alerting a team before a terms update takes effect. No API key, no browser and no proxy are needed.

Run in 30 seconds

  1. Open the Input tab. The prefilled list of 16 well-known vendors is a ready-to-run test.
  2. Click Start. A first dataset with about 30 policy rows is ready in a few minutes.
  3. Open the Policy changes table in the Output tab, or download the dataset as JSON, CSV, Excel or HTML.
  4. Replace the list with your own vendors. Set Compare with to "My previous run" and schedule the Actor to be told only about new changes.

What you get

For every domain the Actor looks for the privacy policy, terms of service and (optionally) cookie policy, first through the footer links of the homepage and then through the usual addresses such as /privacy, /legal/terms and /terms-of-service. Each page is converted to clean text (navigation, footer and scripts removed) and split into sentences. The diff is sentence based, so a reordered paragraph or a punctuation fix is not reported as a change.

FeatureDetail
DocumentsPrivacy policy, terms of service, cookie policy
InputDomains such as stripe.com, or direct policy page URLs
BaselineWayback Machine copy from N days ago (default 365), or your previous run
DiffAdded and removed sentences, counts and change percent
Last changedDate of the first archived snapshot that matches the current text
ThresholdA document counts as changed from 3 changed sentences (adjustable)
Bad pagesNot found, blocked and too short pages return a status and are free
Keys and limitsNo account, no API key, no proxy

Pricing

Price per 1,000 policies checked (Free plan)$11.00
Start fee$0.00005 per run
Plan discountsBronze $10.20, Silver $8.64, Gold and above $6.62 per 1,000
Charged foronly documents that were read and compared (status ok)
Not chargedpages not found, blocked, too short or failed

Input

  • Websites or policy URLs: domains or direct policy page URLs, one per line.
  • Documents to check: privacy, terms and cookies (default privacy and terms).
  • Compare with: wayback (archived copy) or lastRun (what this Actor saw last time; the first run falls back to Wayback).
  • Baseline age (days): which archived copy to use for the Wayback comparison.
  • Changed threshold (sentences): minimum changed sentences to mark a document as changed.
  • Find the last-changed date: search the archive for when the current text first appeared.
  • Max characters of changed text: limit per row for added and removed text.
  • Max documents: cap for the run.

Example input

{
"urls": ["stripe.com", "slack.com", "https://github.com/site/terms"],
"docTypes": ["privacy", "terms"],
"compareTo": "wayback",
"compareDaysAgo": 365,
"maxItems": 100
}

Output sample

{
"url": "https://stripe.com/privacy",
"site": "stripe.com",
"docType": "privacy",
"status": "ok",
"changed": true,
"addedText": "We may use service providers to help us process your data. ...",
"removedText": "We keep personal data for as long as ...",
"addedCount": 14,
"removedCount": 9,
"changePercent": 8.4,
"lastChanged": "2026-03-12",
"baselineSource": "wayback",
"baselineDate": "2025-11-02",
"title": "Stripe Privacy Policy",
"wordCount": 9120,
"textHash": "4f0c9e1a7b",
"checkedAt": "2026-11-10T20:00:00.000Z",
"note": null
}

Output fields

Every row has these fields. Rows that are not read keep the same keys with empty values and a note.

FieldTypeDescription
urlstringAddress of the privacy policy, terms or cookie policy page that was checked
sitestringDomain the page belongs to, for example stripe.com
docTypestringDocument type: privacy, terms or cookies
statusstringok when the page was read and compared; not_found, blocked, too_short or error rows are returned free of charge
changedbooleantrue when at least the minimum number of sentences were added or removed since the baseline
addedTextstringSentences that are in the current policy but not in the baseline copy, one per line
removedTextstringSentences that were in the baseline copy but are gone from the current policy, one per line
addedCountintegerNumber of added sentences
removedCountintegerNumber of removed sentences
changePercentnumberShare of the policy text that differs from the baseline, in percent
lastChangedstringDate (YYYY-MM-DD) of the first Wayback Machine snapshot that matches the current text: the policy has read the same since then
baselineSourcestringWhat the current text was compared with: wayback (archived copy), lastRun (your previous run) or none
baselineDatestringDate of the baseline copy (YYYY-MM-DD)
titlestringPage title of the policy
wordCountintegerNumber of words in the current policy text
textHashstringHash of the normalized policy text: store it to detect a change without keeping the text
checkedAtstringISO timestamp of the run that produced the row
notestringExplains a row that is not ok or a baseline caveat, for example that the oldest archived copy was used

Use cases

  • Vendor risk management: watch the privacy policy and terms of every SaaS vendor and get the exact sentences that changed.
  • Compliance and legal teams: keep an evidence trail of when a policy changed and what it said before.
  • Procurement and security reviews: check whether a supplier changed data sharing, retention or AI training wording.
  • Agencies and consultants: send clients a monthly report of policy changes across their tool stack.
  • Privacy researchers: measure how often and how much big platforms rewrite their policies.
  • Product teams: track how competitors word data use, cancellation and refund terms.

Run it from code

API (returns the rows directly):

curl -X POST "https://api.apify.com/v2/acts/transparent_meteorite~privacy-policy-change-monitor/run-sync-get-dataset-items?token=YOUR_APIFY_TOKEN" \
-H "Content-Type: application/json" \
-d '{"urls":["stripe.com","slack.com"],"docTypes":["privacy"]}'

Python:

from apify_client import ApifyClient
client = ApifyClient("YOUR_APIFY_TOKEN")
run = client.actor("transparent_meteorite/privacy-policy-change-monitor").call(
run_input={"urls": ["stripe.com", "slack.com", "zoom.us"], "docTypes": ["privacy", "terms"]}
)
for row in client.dataset(run["defaultDatasetId"]).iterate_items():
if row["status"] == "ok" and row["changed"]:
print(row["site"], row["docType"], row["lastChanged"], row["addedCount"], "added")

MCP (Claude, Cursor and other agents): add the Apify MCP server and enable this Actor.

{
"mcpServers": {
"apify": {
"url": "https://mcp.apify.com?tools=transparent_meteorite/privacy-policy-change-monitor",
"headers": { "Authorization": "Bearer YOUR_APIFY_TOKEN" }
}
}
}

Then ask your agent: "Did the privacy policy or terms of service of these vendors change this year?" and paste the list.

Integrations and scheduling

ToolHow to use it
n8nAdd the Apify node, choose Run Actor, set the Actor to transparent_meteorite/privacy-policy-change-monitor, pass {"urls": ["={{ $json.domain }}"], "compareTo": "lastRun"}, then add Get Dataset Items and an IF node on changed to alert only on real changes
Make and ZapierUse the Apify app: Run Actor, then read the dataset items
Slack and emailSchedule weekly, set Compare with to My previous run, and send rows where changed is true
Google SheetsAppend the dataset rows to keep a policy change log

Every run's dataset can also be downloaded as JSON, CSV, Excel, XML or HTML. For a monitor, schedule the Actor weekly or monthly and use the previous-run comparison.

Search terms this Actor covers

privacy policy change monitor, terms of service change tracker, privacy policy diff, terms and conditions monitor, track privacy policy updates, vendor privacy policy monitoring, TOS change detection, compare privacy policy versions, privacy policy history Wayback Machine, policy change alert, third-party vendor policy review, when did the privacy policy change, cookie policy change monitor, legal document change tracker.

Limits

  • Pages that need JavaScript to show the text, or that block automated requests, return a status such as too_short or blocked and are free.
  • The Wayback comparison needs an archived copy. When none exists the row says so in note and the next run with Compare with set to My previous run becomes the baseline.
  • The diff reports sentence changes. It does not interpret whether a change is good or bad for you and is not legal advice.
  • The Actor reads public pages at a polite pace and respects the archive's rate limit.

FAQ

Do I need an API key? No. The Actor reads the public pages and the public Wayback Machine.

How does it know the last-changed date? It checks archived snapshots and finds the first one whose text matches the current policy. The policy has read the same since that date.

Can I get only real changes? Yes. Filter on changed equal to true. Raise the changed threshold to ignore small edits.

How do I monitor over time? Schedule the Actor and set Compare with to My previous run. Each run compares with the text stored by the previous run.

What if a site has no standard policy address? Pass the direct policy URL instead of the domain.

Are the changes sentence exact? Yes. addedText and removedText list the sentences, one per line.

Support

Found a bug or need a field? Open an issue on the Issues tab with the run URL and we will reply there.

Changelog

  • 0.1: first release. Privacy, terms and cookie policy diff against the Wayback Machine or the previous run, with the last-changed date.