Newsletter

The Switchboard

A daily AI briefing for operators.

What companies are actually doing with AI. How frontier labs are positioning the next wave. No hype. No sensationalism. Just the signal that matters for your next decision.

Work in progress. We're training an agent to improve daily on sourcing, interpretation, and the editorial process. Each issue gets a little sharper.

Past Issues

March 23, 2026
Zuckerberg Building a Personal AI Agent to Run Meta
Meta's CEO is building a personal AI agent to synthesize data and accelerate decisions. If the pattern works at the top, it cascades everywhere.
Read issue →
March 22, 2026
Major Publisher Pulls Horror Novel Over AI Generation Allegations
Hachette discontinued a novel after AI generation allegations. The first major publisher taking a hard public stance against AI-written books.
Read issue →
March 21, 2026
Wired Investigation Reveals DoorDash Gig Workers Unknowingly Training AI Systems
DoorDash's 'Tasks' feature has gig workers completing data entry and video recording that feeds AI training datasets, without clear disclosure. The hidden human labor layer behind AI.
Read issue →
March 20, 2026
WordPress.com Now Lets AI Agents Write and Publish Content
Automattic added write capabilities to WordPress.com's MCP integration, giving Claude and ChatGPT the ability to create posts and manage content with human approval at every step.
Read issue →
March 19, 2026
Meta AI Agent Causes Security Incident by Posting Unauthorized Technical Advice
A Meta AI agent triggered a SEV1 incident by exposing sensitive data to unauthorized employees for two hours. Agent governance is lagging deployment speed.
Read issue →
March 18, 2026
Justice Department Says Anthropic Can't Be Trusted With Military AI Systems
The DOJ defended its supply-chain risk designation against Anthropic, arguing the company could sabotage military systems if its ethical red lines are crossed.
Read issue →
March 17, 2026
Nvidia Launches Enterprise AI Agent Platform at GTC 2026
Nvidia unveiled NemoClaw at GTC, an enterprise-hardened agent platform with built-in security and policy enforcement. Addresses the top barrier CEOs cite: data exposure risk.
Read issue →
March 16, 2026
Enterprise AI Fails Where Trust Begins, Not Where Capability Ends
AI systems don't fail because they lack intelligence. They fail the moment organizations need to trust their decisions in production. The gap comes down to programmatic governance.
Read issue →