AI Security
Hosted or self-hosted models for security work: how to pick
Pick a hosted model when data can leave your boundary under a contract you trust, and self-hosted when it cannot. Weigh the other trade-offs here.
Oct 18, 20264 min read
Resources
Supply Chain Security, in plain English.
Deep dives, practical guides, and incident analyses from engineers who build Safeguard. No fluff, no vendor FUD — just what you need to ship secure software.
Filter
Articles
RSS feedAI Security
EU AI Act for VPs of engineering: dates, duties, build list
The EU AI Act phases in from 2025 to 2028. Engineering teams should classify systems, build logging and oversight, and check current dates before planning.
Oct 18, 20264 min read
AI Security
AI coding assistant security checklist for engineering teams
Evaluate an AI coding assistant on data handling, output safety, agent permissions, supply chain and logging. Use this checklist before a wide rollout.
Oct 18, 20264 min read
AI Security
Agent identity and short-lived credentials: a checklist
Give each AI agent its own identity, scope tokens to one resource and let them expire fast. A checklist built from SPIFFE, OAuth and MCP specs.
Oct 18, 20265 min read
AI Security
Why is your AI application exposed, and what do you fix first?
AI apps inherit normal app risks and add new ones. Walk the OWASP Top 10 for LLM applications and fix excessive permissions and output handling first.
Oct 18, 20264 min read
AI Security
Why does machine learning matter in cybersecurity, and where does it fail?
Machine learning spots activity no fixed rule describes, but it can be evaded or poisoned. Learn where it helps, where it fails and how to test it.
Oct 18, 20264 min read
AI Security
Why do AI-native apps break traditional AppSec assumptions?
AI-native apps mix instructions and data and act through tools, so classic input checks and code scans miss risks. See what changes and how to adapt.
Oct 18, 20264 min read
AI Security
Why does AI governance lag behind AI pilots?
Pilots need one sponsor and a week. Governance needs owners, inventory and decisions. Here is why the gap opens and a small set of steps to close it.
Oct 18, 20264 min read
AI Security
Why does agentic AI in your pipeline need a new security approach?
A coding agent in CI reads untrusted text and holds real credentials. Here is why old pipeline controls fall short and which limits to put in place first.
Oct 18, 20264 min read
Stay informed
Weekly insights on software supply chain security, delivered to your inbox.
Self-healing security runs on Safeguard.
Your first fix PR is minutes away.
No sales call required, even your agent can complete the purchase over MCP.