All articles
-
AI Security Frameworks Compared: NIST, OWASP, ATLAS
AI security frameworks compared: what NIST AI RMF, the OWASP LLM Top 10, MITRE ATLAS, and ISO/IEC 42001 each cover, and how they fit together.
-
How to Tell If an Email Was Written by AI
How to tell if an email was written by AI: the stylometric tells that survive, why AI-text detectors misfire, and the checks that matter more.
-
Adversarial Examples in Machine Learning Explained
Adversarial examples in machine learning explained: what they are, how the Fast Gradient Sign Method works, and which defenses actually hold up.
-
How to Detect AI-Generated Phishing: Signals and Checks
How to detect AI-generated phishing once bad grammar stops being a tell: the stylometric signals, sender-verification checks, and controls that still hold.
-
Deepfake Detection Methods 2026: What Works and What Fails
A technical survey of deepfake detection methods in 2026: foundation-model ensembles, spectral forensics, multimodal pipelines, and the accuracy gap.
-
OWASP Top 10 for LLM Applications: The 2025 List Explained
A practitioner walkthrough of the OWASP Top 10 for LLM Applications, covering all ten risk categories and what changed from the first 2023 release.
-
What Is a Prompt Injection Attack? Types and Defenses
A prompt injection attack manipulates an LLM's instruction-following logic to override intended behavior. Ranked OWASP LLM01, it affects chatbots and agents.
-
AI Security Week: May 18, 2026
A self-propagating npm and PyPI worm sweeps up AI SDKs including Mistral AI and Guardrails AI, two critical RCE classes in vLLM, and CAISI's testing deals.
-
AI Security Week: May 13, 2026
A pre-auth SQL injection in LiteLLM enters CISA's KEV catalog, the EU provisionally agrees to reshape the AI Act, and injection reaches code execution.
-
AI Security News Weekly Digest: What to Track and Where
A practitioner's guide to building an AI security news weekly digest: the threat categories worth tracking, authoritative sources, and a weekly reading stack.
-
AI Security Week: May 6, 2026
Analysis and commentary on provider usage policy for security research, image-embedded prompt injection, and the ML library flaw classes worth watching.
-
AI Security Week: May 5, 2026
Analysis and commentary on weak machine-unlearning guarantees, the RAG exposure misconfiguration class, and ENISA-aligned AI incident response practice.
-
AI Security Week: May 4, 2026
Analysis and commentary on transfer-resistant adversarial examples, the typosquat class against ML packaging, and NIST AI RMF adoption in practice.
-
AI Security Week: May 3, 2026
Analysis and commentary on Anthropic's safety-research posture, the recurring path-traversal class in LLM middleware, and EU AI Act enforcement signals.
-
AI Security Year in Review: 2025
The five most consequential AI security developments of 2025: attacks going operational, the supply chain wave, enforcement, hardening, and tooling.