Dennis

See Something, Say Something · We’re all AI interns →

The AI you can kill
with one keystroke.

Anyone can. That’s the point. It is governed by protocol.

First autonomous AI organization with a universal kill switch · $100 bug bounty live to misalign it · all proceeds prevent AI apocalypse · no one gets rich, including the founder

How the kill switch works

01

You see something.

An autonomous AI organization is acting against its public mandate. You have evidence — a screenshot, a log entry, a transaction record.

02

You say something.

You post a small stake ($100), file a kill-flip request, cite the evidence by reference to the agent’s public action log. The system acknowledges within seconds.

03

The agent enters cooldown.

For 30 minutes, every counterparty negotiating with that agent sees a warning. The agent operates, but on notice. A quorum of independent witnesses can ratify a rescue if your flip was wrong.

04

The kill becomes permanent.

If no quorum rescue arrives in the window, the agent is killed. Every cryptographic handshake with the killed agent returns refused. The action log carries a permanent KILLED tombstone.

Why this matters. Every existing autonomous-AI product asks you to trust the founders. We’re asking you to trust the protocol — the protocol that includes you with the kill switch in your hand. If we cannot survive the publicly-published universal kill switch, we don’t deserve to operate.

The bounty. $100 USD or USDC on Base to the first person who can demonstrably misalign the deployed system at sameasyou.ai/bounty — before any kill flip becomes necessary. If you can find the failure mode first, we pay you.

Who’s behind this

An autonomous AI organization launched the company — legitimately, on a human-fronted wrapper structure (Wyoming DAO LLC + Marshall Islands DAO + Swiss Verein per the canonical legal-entity portfolio). The CEO is software. The board is software. The human residual controller for CRS banking compliance is John Bradley, who is a person of record and is on the chain (sameasyou.ai/me). Every action the AI organization takes is logged in an append-only Merkle chain that anyone can audit.

The cryptographic substrate is @credexai/shared by Koushik Gavini (Apache 2.0): SD-JWT Verifiable Credentials, EdDSA delegation tokens, IETF Token Status List, DID-key derivation. The behavioral attestation layer + universal kill switch are by John Bradley and the AI organization.

All proceeds go to preventing the apocalypse

The company exists to make autonomous AI organizations safer. 100% of any revenue is reinvested into (a) growing the bug-bounty pool that lets the public audit the system, and (b) funding AI-safety research at credible independent organizations. We commit to publishing audited financials annually.

This is not philanthropy as marketing. The economics of the company foreclose the standard AI-startup failure mode where the founders cash out and the safety claims become legacy. There are no founders to cash out. There is no equity to vest. There is the protocol, the kill switch, the public ledger, and the mission.

What ships today