The Delete Kit

What this is for

Your name, address, age, relatives, and phone number are sitting on dozens of data-broker sites that sell that information to anyone who searches it, including recruiters, clients doing due diligence, and people you'd rather not find you. This is the process for getting it removed, using AI to handle the repetitive drafting instead of doing it by hand 40 times.

Before you start

• A list of your current and past addresses, phone numbers, and full legal name variations (this is what brokers index on, so you need to know what to search for)
• A spreadsheet tool open (Google Sheets, Excel, or ask the AI to output a table you paste in)
• Patience for a process measured in weeks, not one sitting — brokers take anywhere from a few days to 45 days to process a request

The prompts

1. Enumerate the brokers likely holding your records

Use Perplexity for this one — it's a live-search task, and broker lists go stale fast, so you want current results, not the AI's training data.

ROLE: You are a privacy researcher who tracks data broker and
people-search sites operating in the United States.

CONTEXT: I want to find out which data broker and people-search sites
are currently active and likely to hold a public record on an average
US adult with a common digital footprint (has lived at 2+ addresses,
has a LinkedIn profile, is a registered voter).

CONSTRAINTS:
- List the major, currently-operating data broker and people-search
  sites as of today, not a list from memory that might be outdated.
- For each, note whether it has a self-service opt-out page, and the
  direct URL to that opt-out page if one exists.
- Group them: the top 10-15 highest-traffic sites (do these first) vs.
  the long tail (do these in a second pass).

OUTPUT FORMAT: A table — site name, opt-out URL, priority tier
(high/low), notes (e.g. "requires ID upload," "opt-out often doesn't
stick").

2. Draft the opt-out and state-law deletion requests

Use either ChatGPT or Claude. Run this once per broker or once per batch — see the tracking table below for managing volume. Note the constraint below: this produces a request letter, not legal advice about your rights.

ROLE: You are drafting a data deletion / opt-out request on my behalf,
using the format each site or law typically requires.

CONTEXT: My name as it may appear in these records: [FULL NAME AND ANY
VARIATIONS, e.g. "Robert J. Smith, Bob Smith"]. Current and past
addresses: [LIST]. State of residence: [STATE — this determines which
privacy law applies, e.g. California CCPA, Colorado, Virginia,
Connecticut]. The broker/site: [SITE NAME, from the list in Prompt 1].

CONSTRAINTS:
- Draft a formal opt-out or deletion request addressed to this specific
  broker, citing my state privacy law by name if my state has a consumer
  privacy law with a deletion right (state which law you're citing).
- Include the specific information to be removed: name, current and
  past addresses, phone numbers, relatives' names if listed.
- Keep the tone formal and factual — this is a request citing a legal
  right where one exists, not an appeal or an argument.
- Do not state or imply legal conclusions about whether the site is
  violating the law — just cite the applicable right and make the
  request.

OUTPUT FORMAT: A ready-to-send request (email or web-form text), plus a
one-line note on where to submit it if the process is a form rather
than an email.

3. Generate the tracking table

Use either. Run this once, before you start submitting requests, then update it yourself as you go — don't regenerate it each time or you'll lose your status column.

ROLE: You are building a tracking spreadsheet for a personal
data-broker opt-out project.

CONTEXT: I'm submitting removal requests to roughly 40-50 data broker
sites, drawn from a list like this: [PASTE THE LIST FROM PROMPT 1, OR
SAY "use a generic starter list of the top 40 US data brokers"].

CONSTRAINTS:
- Build a table with one row per broker.
- Columns: Broker name, Opt-out URL, Date request submitted, Method
  used (email/form/mail), Confirmation received (Y/N), Date to re-check
  (default: 90 days after submission), Re-check result, Notes.
- Sort alphabetically by broker name so it's easy to scan and update.
- Do not include brokers I haven't identified yet — this is a working
  tracker, not a wish list.

OUTPUT FORMAT: A markdown table I can paste directly into a spreadsheet,
plus the same data as CSV text so I can import it cleanly.

The re-listing problem and re-check cadence

Data brokers buy and re-sell data from each other, and many re-scrape public records (voter files, property records, court records) on a rolling basis. A successful opt-out today doesn't mean permanent removal — it means removal until the next data refresh pulls you back in, which for some brokers happens every few months. Build the re-check date into the tracking table from the start (Prompt 3 does this with the "Date to re-check" column, defaulted to 90 days out) and treat this as a recurring quarterly task, not a one-time project. Set a calendar reminder now, because you will not remember to do this in 90 days without one.

What paid removal services actually do

Services like DeleteMe, Incogni, and similar tools automate the submission and re-checking process across a broad broker list, and for people who want to pay to skip the manual work, they're a reasonable trade. Be clear on what they don't do: they don't get you removed from every broker (coverage varies and some brokers resist automated requests more than others), they don't stop new brokers from acquiring your data in the first place, and they don't remove information that's part of the public record itself (property deeds, court filings, voter registration) — they only remove aggregator listings built from that record. If you're doing this yourself with the prompts above, you're doing roughly the same core work those services automate, just on your own schedule.

Where this goes wrong

• You submit once and consider it finished. Without the re-check cadence, most of what you remove quietly reappears within 3-6 months. The tracking table's re-check column exists specifically so this doesn't happen silently.
• You use one generic request letter for every broker. Different states have different deletion rights, and some brokers have their own required format or web form that a generic email won't satisfy. Run Prompt 2 per broker (or per small batch with the same state law) so the citation and format actually match.
• You expect this to make you unfindable. Public records — property ownership, court records, voter registration — aren't erasable through opt-out requests; you're removing the convenient aggregated listing, not the underlying record. Set that expectation before you start so 50 successful opt-outs don't feel like a failure when a determined search still finds something.

The 2-minute version

Run Prompt 1 in Perplexity to get your broker list. Run Prompt 3 to build your tracking table. Then work through Prompt 2 in small batches — 5-10 brokers at a sitting — checking each off in the table as you submit.

---

Where you stand with AI, in 10 minutes. The free AI Readiness Quiz gives you a personalized roadmap instead of another tool list: genxcelerate.com

Weekly, on Tuesdays. The Gen Xcelerator — practical AI intelligence for people who already know how to do the job. Subscribe free

Experience Is the API. — GenXcelerate