Information Taser — AI Agent Off Switch

Foresight, filed early. Not a retrofit.

The AI-insurance framework around the Information Taser was authored in 2017+, before the 2024+ AI-safety conversation became mainstream. The package contains the concept, the public-facing AI-insurance HTML page, the supporting RAG-style indexing scripts, and drafted formal outreach to the FBI, the White House, and the Scammer Payback program. Acquisition is copy-ownership of the assembled artifact — the work itself, on the record, dated.

What This Package Contains

The Information Taser is a concept artifact with attached operational and outreach materials. It is the productised expression of an idea the inventor calls AI Insurance — that AI agents which take real-world actions need a public, named, legally-recognised off switch, and that the time to have written that switch is before the agents are running, not after.

The package on disk holds three kinds of content:

Maturity and Scope — Stated Honestly

Per the project's MANIFEST, the Information Taser is at L1 (Research / Outreach in Progress). That is the honest label. The package is concept-grade with formal outreach materials drafted; it is not a deployed enforcement product, and it is not a regulatory instrument. What it is, exactly, is a copy-ownership artifact representing the inventor's AI-insurance position — written early, kept on the record, available for acquisition by a buyer who wants the assembled work as an artifact.

USPTO patent status: no specific application is editorially linked to this project at the time of writing. The companion project 43 — Redaction Pen · Protect and Defend (covering applications 17/687,656, 18/650,137, 19/031,884, 19/177,547, 63/552,008) is thematically related; both touch on truth, authorship, and public discourse.

How it's made

The Information Taser is assembled, not coded. Its center is an essay — Respect the Tool — written by the inventor on AI-as-tool ethics. Around the essay sit the public-facing AI-insurance HTML page (which carries the same argument in storefront form), the RAG-style indexing scripts (which let a reader query the inventor's body of work for prior-art coincidences), and the formal outreach drafts (which are the written communications the inventor would send if a relevant agency engaged).

Every artifact in the package is a file the inventor authored. None of it was generated to order for a buyer. The package's value is precisely that it pre-exists the moment a buyer might want it.

Why I made it

In 2017 I started thinking carefully about what would happen when software systems began acting in the world without continuous human approval — not because I expected the agents would be malicious, but because I expected the legal, regulatory, and operational vocabulary around them would lag the technology by a decade or more. I wrote the AI-insurance position as a way of getting on record — before the conversation existed in public — that an off switch belongs in the design, not the patch.

The Information Taser is the named, productised version of that position. The outreach materials are the path I would take to bring it forward in a serious conversation. I built it now so that, by the time a regulator or an agency wants it, it is already written.

What it can do

An acquirer takes possession of the complete AI-insurance artifact: the concept essay, the public-facing HTML page, the RAG-style index over the inventor's body of work, the coincidence-finding analysis script, three drafted formal outreach letters (FBI, White House, Scammer Payback), and the cross-cutting design notes that connect this project to the rest of the portfolio.

The package can serve three concrete functions for a buyer:

• Prior-art anchor. A dated, on-disk artifact establishing the inventor's 2017+ position on AI-agent off-switch / AI-insurance design.
• Narrative seed. The Respect the Tool essay and the AI-insurance HTML page are publication-ready in the inventor's voice.
• Outreach kit. If the acquirer is positioned to forward the FBI / White House / Scammer Payback drafts (or comparable agencies in their jurisdiction), the drafts are written and on file.

Why it's a fact

The claims above are checkable against the source record:

• The package files exist on disk at D:\special\29-information-taser\ with file timestamps that pre-date the 2024+ AI-safety conversation. The dates can be inspected.
• The Respect the Tool essay exists in two formats (HTML + text) within the package; either can be read end-to-end before purchase.
• The three outreach drafts are present as EMAIL_FBI_OUTREACH.md, EMAIL_WHITE_HOUSE_OUTREACH.md, and EMAIL_SCAMMER_PAYBACK_OUTREACH.md. Their contents are inspectable.
• The companion Project 43 (Redaction Pen) and its associated USPTO applications — 17/687,656, 18/650,137, 19/031,884, 19/177,547, 63/552,008 — are independently filed instruments that anchor the broader truth-and-authorship work this project sits inside.
• The maturity label L1 (Research / Outreach in Progress) is the inventor's own assessment in the project's MANIFEST. Not deployed. Not regulatory. Concept + outreach. Stated honestly.

License Terms — What's Granted, What Isn't

The acquisition grants the buyer permission to make, build, and copy the deliverable. It does not transfer the underlying intellectual property:

• Granted with the acquisition: permission to make use of the assembled artifact for the buyer's private study, archival, prior-art positioning, and outreach activities; permission to build the materials into the buyer's own AI-policy work or onward outreach to comparable agencies; permission to make copies of the package for the buyer's organisational use.
• Not transferred with the acquisition: the patents and patent applications referenced through the companion projects (Project 43 USPTO 17/687,656; 18/650,137; 19/031,884; 19/177,547; 63/552,008), the inventor's broader portfolio book references, trademarks, copyrights, or any rights to license or assign the IP onward. The intellectual property remains held by Christopher Gabriel Brown.
• The buyer's permission is to use the artifact, not to own the rights behind the artifact.

This framing applies uniformly across the inventor's portfolio. Buyers seeking IP assignment rather than make/build/copy permission should contact the inventor directly — that is a separate negotiation outside the standard storefront acquisition.

Foresight is dated, or it isn't foresight.

What gives this package value is not what it might one day become — it is the date stamps on the files inside it. The argument for an AI off-switch was on the record before the agents were. That is the entire point of acquiring it.