AI Security
In-depth guides and analysis on ai security from the Safeguard engineering team.
786 articles
Supply Chain Attacks Targeting AI/ML Pipelines
AI and ML pipelines introduce unique supply chain risks -- from poisoned training data to compromised model registries. Here is what attackers are targeting and how to defend.
Open-Weight Model Sandboxing Patterns
Running an open-weight model inside an enterprise perimeter seems safer than calling a hosted API. It is, and it isn't. The sandboxing patterns that actually produce the safety properties.
Security in AI Systems: What Actually Changes
Security in AI systems isn't a wholly new discipline, but prompt injection, training data provenance, and model supply chains introduce risks traditional AppSec tooling wasn't built to catch.
Introducing Agentic Development Security (ADS)
As AI agents now author up to half of production commits, Safeguard introduces Agentic Development Security (ADS) — a new framework for securing autonomous coding.
Safeguard Now Supports Every Major AI Model Family for Zero-Day Discovery: Anthropic, OpenAI, Gemini, Microsoft, Meta, and Your Own Models
You should not have to choose between your organization's AI strategy and your security platform. Safeguard's agentic zero-day discovery and remediation pipeline now works on Anthropic Claude Fable 5, OpenAI GPT, Google Gemini, Microsoft Phi, Meta Llama, Safeguard native models, and privately hosted custom models — all running as first-class agents in the same Multi-Agent TAOR Deep Think AI Engine.
Anthropic Claude Mythos Releases Tomorrow: Capabilities, Benchmarks, and What Security Teams Must Do Now
Anthropic's Claude Mythos model goes public on June 10, 2026 — a frontier AI that scored 97.6% on the Math Olympiad, completed expert-level hacking tasks at 73% success, and found 271 vulnerabilities in Firefox 150. Here is everything security teams need to know before it lands, and how Safeguard already supports Mythos zero-day discovery natively.
Claude Fable 5: Anthropic's Most Capable Public Model Is Here — Benchmarks, Capabilities, and What It Means for Security
Anthropic just released Claude Fable 5, its most capable publicly available model and the first Mythos-class AI open to everyone. 80.3% on SWE-Bench Pro, 88% on Terminal-Bench 2.1, state-of-the-art across software engineering, vision, and scientific research. Safeguard has already integrated Fable 5 natively — here is everything you need to know.
GPT-5 Launch: Reading the System Card for Supply-Chain Risk
GPT-5 shipped August 13, 2025 under OpenAI's Preparedness Framework v2. Here's what the system card tells security teams about deployment risk.
Data poisoning attacks against LLMs
A $60 domain purchase or 250 documents can backdoor an LLM. Here's how data poisoning attacks work, real cases, and how to defend against them.
Securing RAG pipelines against injection attacks
RAG pipelines feed untrusted retrieved content straight into model context. Real breaches like EchoLeak show what happens when nothing checks it.
Type-level security: the future of secure AI code generation
45% of AI-generated code fails basic security tests. Here's why type systems catch what code review misses, and how to enforce type-level security on AI-authored diffs.
Snyk VulnBench: benchmarking LLMs on repeat vulnerability discovery
Snyk's VulnBench JS 1.0 ran 300 repeated LLM scans and found half of non-reference findings vanish on rerun—raising the bar for AI security tooling.
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.