>

Weekly Briefing

AI Security Weekly

The essential weekly briefing for enterprise AI security. Every signal, every shift, every decision point — distilled into one actionable report.

200+ Companies Tracked
10 Market Categories
Weekly Delivery Cadence

Join security leaders across enterprise, defense, and government. One briefing, five minutes, every week.

Latest Issue

Issue #12 — June 2026

Current Issue

AI Security Weekly

Issue #12 — June 2026

01

The Principal-Agent Question Reaches the Cyber Form

Flowise CVE-2026-40933 (1-click stdio MCP RCE) and LibreChat CVE-2026-44653 (MCP secret exposure, fixed in 0.8.4) turn the named-insured question from a thought experiment into a claims question. The connector substrate that delegates agent authority is now the substrate adversaries are targeting — and current cyber and E&O forms were drafted before any of those words meant what they now mean.

02

Runtime Containment as Underwriting Evidence

Microsoft Build 2026 published Execution Container SDK, Foundry runtime DLP, and Defender AI model scanning in coordinated rollout. Together they produce the audit substrate an underwriter can score against and a regulator can read as conformity evidence — the first runtime stack designed for cross-audience evidence. Anthropic’s CVD dashboard at 1,596 vulns / 281 projects is now part of the supply-side file.

03

Five Moves Before the Underwriter Asks

The Article 50 consultation closed June 3 with eight weeks to the August 2 effective date. What every high-risk deployer should be doing now: inventory MCP connectors, containerize agent execution, treat model-provider CVD as supply-side file, close the Article 50 implementation loop, and document the principal-agent chain. Market Index W23: 37.7 (flat against W21).

Read Issue #12 →

Coming Next

On the Horizon

Issue #13 Mon, Jun 15, 2026

The Software Bill of Materials for AI Stacks

W20 surfaced a transitive-dependency aggregation pattern the npm ecosystem can no longer absorb quietly. What happens when that pattern repeats at AI-stack scale — and why SBOMs for model weights, datasets, and inference dependencies become the next underwriting precondition.

Get Notified →
Issue #14 Mon, Jun 22, 2026

The Underwriting File for High-Risk AI

Once Article 50 documentation exists, what does a carrier actually want to see? The composition of the underwriting file — conformity narrative, post-market monitoring outputs, incident classifications, defensive-AI disclosure — and how AIRS maps to the file a syndicate will price against.

Get Notified →
Issue #15 Mon, Jun 29, 2026

The Reinsurance Treaty Cycle Meets AI

January 1 renewals are the first treaty cycle to fully price agent-runtime risk. Early signals from June and July treaty placements set the W26+ tone — what the cedent disclosures are starting to require, and which lines of evidence the treaty market is pricing first.

Get Notified →

What You Get

What Each Issue Covers

Market Signals

Funding rounds, M&A activity, product launches, and strategic partnerships — tracked in real time across 200+ companies.

Competitive Analysis

Category movements, vendor positioning shifts, and emerging competitive dynamics that reshape the landscape week by week.

Regulatory Watch

Policy shifts, compliance deadlines, and government action across the EU AI Act, US executive orders, and global AI governance frameworks.

Strategic Commentary

Expert analysis on what the signals mean for your organization — contextualized insights that connect market movements to decision points.

Never Miss a Briefing

Join security leaders across enterprise, defense, and government organizations. One briefing, five minutes, every week.

Free weekly intelligence  ·  Five-minute read  ·  Unsubscribe anytime