Protective intelligence
ObscureIQ / Intelligence deskIndependent analysis. Protective judgment.
← Back to the libraryRecord / CB-245-A
Annex / CB-245-A

The custody map and the never list

The intelligence

A custody map template and a never list, companion to CB-245, with the checks an administrator can run inside the tenant alone, a letter for the provider and a guide to grading the replies. It lets a family office describe what an approved AI platform holds, who can reach each part, and what must never be entered at all. For administrators and counsel at procurement, contract renewal, or whenever a new connector or model is added.

COMMAND intelligence

Unlock this Command Brief.

Get this assessment, the full subscription collection, and new intelligence as it is published.

Introductory rate · through 30 November
$75/month

For your first year, then $100/month. Or $750 for the first year, then $1,000 annually.

  • The CORGI Morning Brief, each morning
  • This brief and the subscription collection, 100+ briefs
  • New Command Briefs as they are released
  • Principal and Analyst briefs, as PDF
Subscribe to COMMAND

Secure checkout through Substack. Substack asks for your email first.

Full subscription details

Topics

Hosting & InfrastructureData MinimizationOperational Hygiene
Follow the thread

Cross-referenced intelligence.

CB-245subscription
Command Brief

The sanctioned counterparty: consolidation moves the exposure

They Kept the Promise and Kept the Data

The usual fix for staff using AI in personal accounts is one approved platform on enterprise terms, and in 2025 a federal court ordered an AI provider to preserve conversations its users had deleted. This brief asks what consolidation actually changes about a principal's exposure and how much a no-training pledge covers. Built from vendor telemetry, breach-cost research, that court record and observation of household use. For family offices and counsel choosing and governing an approved platform.

Household & StaffHosting & InfrastructureData Minimization
View record
CB-238subscription
Command Brief

Consent phishing: the attack you approve yourself

Nothing Was Forged

On 1 September 2026 the FBI warned of a campaign that messages prominent people and their families and induces them to approve a malicious application against a real cloud account. Through composite reconstructions, this brief asks what the approach looks like on screen, how it differs from other phishing, and what recovery must involve. Built from the Bureau's alert and independent research on malicious app registrations. For family offices and whoever administers household accounts.

Social EngineeringEmail SecurityHousehold & Staff
View record
CB-219subscription
Command Brief

Counterparty Agent Exposure: the ChatGPT Messages Plugin

Your Texts, Read by Someone Else's AI

In August 2026 a ChatGPT plugin began letting Mac users read, search, summarize and send Messages through the assistant. This brief looks at the release from the side of a principal who never installs it but whose counsel, assistants and family might, and asks what controls exist for others in a thread. Built from vendor and platform documentation, press reporting and 2026 privilege rulings. For family offices and counsel managing the people who hold a principal's correspondence.

Secure MessagingHousehold & StaffMobile & Devices
View record