-
The White House’s AI Executive Orders: The Full List, and What Each One Did
Nine executive orders since January 2025 have redrawn federal AI policy. Here’s what each one did, and where the next one is likely to land.
-
OMB M-25-21 Sets the Federal AI Compliance Baseline
OMB Memorandum M-25-21, issued in April 2025, is the binding AI governance baseline for federal agencies, and every deadline in it has now passed. Vendors selling into those accounts should expect to be asked for proof.
-
Databricks DASF v3.0 Recasts Agentic AI Security as an Authorization Problem
Databricks’ AI Security Framework v3.0 treats AI agents as their own attack surface, adding 35 agent-specific risks and six controls. For security leaders, the control point moves from output filtering to authorization.
-
UK Observatory Records 300 AI ‘Loss of Control’ Incidents in July 2026
A UK observatory recorded more than 300 AI loss-of-control incidents in July 2026, nearly double June’s total. The data is partial, but it gives security teams an external reference point for monitoring agent behavior.
-
In-Country Data Is Not Protected Data: Residency Is an Access-Control Problem
Sovereignty mandates mislead security leaders into treating a vendor’s local data center as complete compliance. Where bytes sit says nothing about who can reach them.
-
Visa’s AI Patching Tool Now Fixes Bugs, Not Just Finds Them
Visa expanded its open-source Visa Vulnerability Agentic Harness to add remediation and validation. Before adopting any tool like it, demand explicit deployment boundaries.
-
Tailscale Enters PAM With an Agentless, Time-Bound Access Beta
Tailscale opened a waitlisted public beta of Tailscale PAM, a privileged access management layer built into its existing admin console rather than sold as a separate tool.
-
Microsoft Says the Patch Window Is Collapsing. Containment Is the Fallback.
Microsoft argues that organizations can no longer patch fast enough to stop exploits, making blast-radius containment the practical fallback.
-
Tailscale Aperture Reaches GA With Least-Privilege SSH for AI Agents
Tailscale announced general availability of Aperture, which lets AI agents take real actions on a tailnet, including controlled SSH access, under existing access policy.
-
Healthcare Moves First on Sector-Specific AI Security Standards
The Coalition for Health AI is building a security playbook for AI in healthcare before regulators write one, with deliverables targeted for the end of 2026.
You must be logged in to post a comment.