AI Security
In-depth guides and analysis on ai security from the Safeguard engineering team.
678 articles
Continuous Eval & Release Gating: Griffin AI vs Mythos
Evals that run once are marketing. Evals that run on every build are infrastructure. Griffin AI runs the harness on every change; Mythos does not describe one.
Race Condition Detection: Griffin AI vs Mythos
Race conditions are the hardest class of vulnerabilities for static analysis. Specific architectural capabilities separate tools that find them from tools that claim to.
LLM Selection For Security Workflows
Picking a model for a security workflow is not the same as picking one for a chatbot. Here are the criteria that actually matter and how to weigh them.
Open-Source LLM Supply Chain Incidents 2026
Open-source LLM ecosystems hit a turning point in 2026 as supply chain incidents — backdoored fine-tunes, compromised weights, malicious adapter packages — moved from rare to recurring.
Enterprise AI Data Residency Requirements, 2026
Data residency for AI workloads has moved from nice-to-have to contractually required. The shape of the requirement is specific and worth knowing before procurement.
From CVE To PR: The Full Remediation Pipeline
A complete walkthrough of the modern remediation pipeline, from advisory ingestion through merged and deployed fix, with every stage that actually matters.
False Positive Cost: Griffin AI vs Mythos
A false positive is not free. It costs engineer attention, trust in the tool, and eventually the security programme's credibility. We price the difference.
Injection Path Detection: Griffin AI vs Mythos
Injection vulnerabilities are not really about the sink. They are about the path from untrusted input to the sink. The path is where Griffin AI and Mythos-class tools diverge.
Griffin AI vs Open Weights: On-Prem Tradeoffs
Open-weight models let you run everything locally. The tradeoff is quality, cost, and operational overhead. Griffin AI provides a different answer to the same on-prem need.
Fine-Tuning Poisoning Detection for Supply Chains
Fine-tuning inherits every problem of the base model and adds dataset provenance as a new one. Here is how detection actually works in practice.
RAG Pipeline Security Controls in 2026
Retrieval-augmented generation pipelines have become a primary breach vector for LLM products. The controls that contain the risk without breaking the use case.
LLM-As-Judge Pitfalls In Security Evals
Using an LLM to score another LLM's output is expedient and dangerous. The judge has its own biases — ones that affect security evaluations specifically.