Tagged:

8 posts

• Mechanistic Interpretability as a Security Tool: Detecting Backdoors and Hidden Behaviors in AI Models

  Circuits, sparse autoencoders, and activation steering let security teams detect backdoors and hidden model behaviors that red-teaming cannot find.

  Jul 11, 2026 mechanistic-interpretability backdoor-detection model-security ai-safety pre-deployment-inspection sparse-autoencoders

• Model Hub Supply Chain Attacks: Malicious Models, Tokenizer Exploits, and Typosquatting on Hugging Face

  Downloading an open-weight model from a public hub is not a read-only operation. Custom tokenizer classes and auto_map configs execute arbitrary Python when trust_remote_code is set, LoRA adapters can trojanize safe base models, and typosquatted namespaces are a documented distribution vector. Here's the threat model practitioners need before they run from_pretrained().

  Jul 9, 2026 supply-chain model-security ai-security security data-poisoning pickling fine-tuning threat-modeling

• AI Incident Response: A Practitioner's Playbook for When Your AI System Is Compromised

  Detection signals, containment options, evidence preservation, and recovery procedures for AI-specific security incidents — the operational complement to attack coverage. NIST IR lifecycle applied to prompt injection, model backdoors, data poisoning, and adversarial input attacks.

  Jul 7, 2026 incident-response prompt-injection model-security mlops defense playbook

• ML Model Provenance: Signing, SBOMs, and Verifying the AI You Deploy Before It Runs

  You wouldn't deploy software without checksums and signatures. But most organizations download model weights and run them without any provenance verification at all. This post covers the practical mechanics of model signing, ML SBOMs, and the emerging infrastructure for verifying a model's origins before it touches production.

  Jul 6, 2026 supply-chain-security model-security defense-patterns content-provenance

• Malicious AI Model Files: Pickle Exploits and Arbitrary Code Execution on Model Load

  Downloading a model file and calling torch.load() is a potential code execution event. This post explains the pickle mechanics that make it so, the real-world exploits already found in the wild, how HuggingFace has responded, and what safe alternatives actually prevent the attack.

  Jul 5, 2026 supply-chain-security model-security deserialization pickle pytorch safetensors defense-patterns

• Backdoor Attacks in Foundation Models: Sleeper Triggers That Survive Fine-Tuning

  Pre-trained LLMs can be trojaned at the foundation stage, with adversarial triggers embedded in weights that persist through downstream fine-tuning and RLHF safety training. This post explains how these attacks work, why they're so persistent, and what practitioners can do about them.

  Jul 3, 2026 backdoor-attacks supply-chain-security adversarial-ml fine-tuning model-security trojan

• Machine Unlearning Security: When Forgetting Training Data Creates New Vulnerabilities

  Regulators are demanding AI systems forget specific training data. Attackers are figuring out how to weaponize that forgetting — through verification gaps, induced forgetting abuse, and unlearning as a path to alignment degradation.

  Jul 2, 2026 machine-unlearning privacy gdpr safety-alignment adversarial-ml model-security membership-inference

• Fine-Tuning Trojans: Injecting Backdoors Through the Model Training Pipeline

  How malicious training data, tampered datasets, and compromised fine-tuning APIs plant backdoored behavior in legitimate base models — and what defenders can do.

  Jun 28, 2026 supply-chain backdoor fine-tuning model-security dataset-poisoning adversarial-ml