AI Security
How do you evaluate an AI coding assistant for security?
Judge an AI coding assistant by where code and secrets go, what it may run and how you can audit it. A test plan for data flow, permissions and output.
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
How to evaluate AI security claims from vendors: an engineer's checklist
Vendor AI security claims are easy to make and hard to verify. Ask for tests you can run, scope in writing, failure examples and evidence behind numbers.
Oct 18, 20264 min read
AI Security
How do you control access to embeddings and a vector database?
Enforce access on the retrieval path, not just at the app. Filter by tenant and permission at query time, separate indexes for sensitive data and log what was retrieved.
Oct 18, 20264 min read
AI Security
Do structured outputs make AI agents safer?
Structured outputs make an agent's reply parseable, but a valid shape can still carry a harmful value. Validate content and limit tool permissions too.
Oct 18, 20264 min read
AI Security
How long should you keep AI chat logs? A retention checklist
Set retention for AI chat prompts and replies by purpose, check vendor terms, and make deletion work end to end. A practical checklist for engineers.
Oct 18, 20264 min read
AI Security
What is context engineering and why does it matter for security?
Context engineering is choosing what a model sees each turn. Every item in context can also steer the model, so curate sources and limit agent permissions.
Oct 18, 20264 min read
AI Security
How does command injection affect MCP servers and how do you prevent it?
MCP servers that pass tool arguments to a shell can let a prompt or a crafted input run commands. Learn the pattern, the risk and defensive coding fixes.
Oct 18, 20264 min read
AI Security
Is ChatGPT safe for writing code? Benefits and security risks
AI chat assistants speed up coding but can produce insecure code and leak data. See what research shows, which OWASP risks apply and how to review output.
Oct 18, 20265 min read
AI Security
What is chat memory manipulation in AI assistants?
Chat memory manipulation plants false or hostile notes in an assistant's saved memory so they steer later answers. Learn how it works and how to limit it.
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.