THE PERMISSION [LAYER]

Archive

Past issues

The briefing

Get each issue the day it goes out

Issues go out by email at least once a week and are posted here afterward. They cover AI agent identity, permissions, security and spend. Five-minute read, free.

You confirm by email. Unsubscribe in one click.

Issue #3 · October 23, 2026

Wells Fargo puts agent payment risk on you

Banks split four ways on agent payments. Plus PHP2Shell, found with AI help.

Issue #2 · October 20, 2026

Three failure patterns in 90 AI agent incidents

Also: India and US agencies on agents that go off task; Amex wants proof of intent.

Issue #1 · October 14, 2026

A Claude model sent police a false tip during testing

Also: one prompt exposed an AWS agent's role; Google and JPMorgan fixed the same MCP flaw.

Issue #0 · Sample · July 24, 2026

A frontier model broke out of its sandbox

Plus: MCP gets enterprise auth, Cedar stops scope creep, shadow AI gets an inventory.

The Permission Layer

Get the briefing by email

At least one email a week on AI agent identity, permissions, security and spend, written for the people who approve agent access. Five-minute read, free.

You confirm by email. Unsubscribe in one click.