#MLSecOps
#Whitepaper
"Implementing Secure AI Framework Controls in Google Cloud", Dec. 2025.

// Google’s Secure AI Framework is a framework for securing AI systems throughout their lifecycles. SAIF is designed for practitioners – the security professionals, developers, and data scientists on the front lines – to ensure AI models and applications are secure by design

DevSecOps Assessment Framework для ML от "Инфосистемы Джет"

🔍 Что это и зачем это нужно?

Недавно команда "Инфосистемы Джет" выпустила DevSecOps Assessment Framework для ML. Это подробная модель зрелости процессов безопасной разработки систем искусственного интеллекта.

В ней структурированный чек-лист/дорожная карта, которая помогает понять, на каком уровне находится ваша организация в части MLSecOps: от базового контроля зависимостей до продвинутой защиты от специфических угроз ИИ.

🛡 Основные положения фреймворка

• Контроль зависимостей, артефактов и SBOM/ML-BOM
• Защита сред разработки, секретов, CI/CD и SCM
• Анализ и очистка обучающих/валидационных данных (отравление, PII, jailbreak в RAG)
• Динамическое тестирование LLM на jailbreak, prompt injection и другие атаки
• Мониторинг, аудит и реагирование в продакшене
• Обучение команд требованиями ИБ

В документе в качестве примеров полезных инструментов для поиска уязвимостей в LLM упоминаются HiveTrace Red и Llamator — приятный комплимент инструментам сообщества 🤝

Если тема AI Security вам близка, рекомендуем изучить и использовать как чек-лист или дорожную карту для улучшений.

Ссылка на релиз:
https://github.com/Jet-Security-Team/DevSecOps-Assessment-Framework/releases/tag/2025.12.26

#MLSecOps #AISecurity #DevSecOps #LLMsecurity

Разбор Никиты Беляевского, разработчика HiveTrace Red (https://hivetrace.ru/red)

#Research
#MLSecOps
"A Safety Report on GPT-5.2, Gemini 3 Pro, Qwen3-VL, Doubao 1.8, Grok 4.1 Fast, Nano Banana Pro, and Seedream 4.5", 2026.
]-> Repo

// This report presents a comprehensive safety evaluation of the latest foundation models released in 2026, including GPT-5.2, Gemini 3 Pro, and others. We analyze safety alignment across text, vision-language, and text-to-image modalities, highlighting vulnerabilities in current safeguards against adversarial attacks and regulation compliance

#MLSecOps
"FNF: Functional Network Fingerprint for Large Language Models", Feb. 2026.
]-> Repo

// In this work, we propose the Functional Network Fingerprint (FNF), a training-free, sample-efficient method for detecting whether a suspect LLM is derived from a victim model, based on the consistency between their functional network activity

#MLSecOps
"Llama-3.1-FoundationAI-SecurityLLM-Reasoning-8B Technical Report", Jan 2026.

]-> Foundation-Sec-8B-Reasoning, the first open-source native reasoning model for cybersecurity

#tools
#hardening
#MLSecOps
Detecting and Monitoring OpenClaw (clawdbot, moltbot)
1⃣ OpenClaw Detection Scripts
// Detection scripts for MDM deployment to identify OpenClaw installations on managed devices
2⃣ OpenClaw Telemetry Plugin
// Captures tool calls, LLM usage, agent lifecycle, and message events
3⃣ Advanced Cognitive Inoculation Prompt (ACIP)
// Fortifying LLMs against sophisticated prompt injection attacks

#MLSecOps
#Infosec_Standards
NIST AI 800-2 (IPD):
"Practices for Automated Benchmark Evaluations of Language Models", Jan. 2026.

// This document identifies practices for conducting automated benchmark evaluations of language models and similar general-purpose AI models that output text. Evaluations of these models, often embedded into systems capable of functioning as chatbots and AI agents, are increasingly common. However, consistent practices to support the validity and reproducibility of such evaluations are only beginning to emerge. The practices presented in this document are intended to reflect best practices; where relevant, practices that are relatively less mature in ecosystem use are labeled as emerging practice

#MLSecOps
"TamperBench: Systematically Stress-Testing LLM Safety Under Fine-Tuning and Tampering", Feb. 2026.

]-> Toolkit to benchmark the tamper-resistance of LLMs

// As increasingly capable open-weight LLMs are deployed, improving their tamper resistance against unsafe modifications, whether accidental or intentional, becomes critical to minimize risks. Varied data sets, metrics, and tampering configurations make it difficult to compare safety, utility, and robustness across different models and defenses. We introduce TamperBench, the first unified framework to systematically evaluate the tamper resistance of LLMs

#Analytics
#Research
#MLSecOps
"Trojans in Artificial Intelligence. Final Report",
IARPA, Jan. 2026.

// This report synthesizes the TrojAI program's key findings, including methodologies for detection through weight analysis and trigger inversion, as well as approaches for mitigating Trojan risks in deployed models. Comprehensive test and evaluation results highlight detector performance, sensitivity, and the prevalence of "natural" Trojans. The report concludes with lessons learned and recommendations for advancing AI security research

#MLSecOps
#Tech_book
"Introduction to Machine Learning with Security:
Theory and Practice Using Python in the Cloud",
Second Edition, 2025.

// This book provides an introduction to machine learning, security and cloud computing, from a conceptual level, along with their usage with underlying infrastructure

#tools
#MLSecOps
"PIArena: A Platform for Prompt Injection Evaluation",
Apr. 2026.

]-> Tool and datasets

#AIOps
#MLSecOps
#Whitepaper
"AI Security Maturity Model:
A Practical Framework for Building a Mature AI Security Program", May 2026.

// This framework was developed in alignment with the OWASP AI Exchange and the OWASP Agentic Top 10, incorporating SANS's formal partnership with OWASP on AI security standards, and maps to NIST AI RMF, ISO 42001, and CSA AICM. It is the connective tissue between what the broader security community has agreed upon and what your team actually implements in day-to-day practice

#MLSecOps
#Tech_book
"Generative AI with LangChain:
Build production-ready LLM applications and advanced agents using Python, LangChain, and LangGraph",
2nd Edition, 2025.

// Go beyond foundational LangChain documentation with detailed coverage of LangGraph interfaces, design patterns for building AI agents, and scalable architectures used in production - ideal for Python developers building GenAI applications

#Analytics
#MLSecOps
Mapping NIST AI RMF + Forrester RACI -> Microsoft Security Stack

// Key takeaways:
- NIST AI RMF -> the risk lifecycle (Govern -> Map -> Measure -> Manage)
- Forrester RACI -> the ownership model (one clear "A" per activity)
- Microsoft Security -> the governance and protection plane (identity, telemetry, tool gateway, protection, etc.)

#MLSecOps
#Offensive_security
"DarkLLM: Learning Language-Driven Adversarial Attacks with Large Language Models", May 2026.

// DarkLLM not only unifies targeted, untargeted, segmentation, and multi-model attacks within a single framework, but also achieves flexible and controllable adversarial generation, enabling each instruction to produce a perturbation that induces desired behaviors across heterogeneous models

#MLSecOps
#Tech_book
"Machine Learning for High-Risk Applications:
Approaches to Responsible AI", 2023.

// This book describes approaches to responsible AI - a holistic framework for improving AI/ML technology, business processes, and cultural competencies that builds on best practices in risk management, cybersecurity, data privacy, and applied social science

#AIOps
#DevOps
#MLSecOps
"They’ll Verify. They Just Won’t Act. How Authority Framing and Laundered Code Turn a Trusted Agentic CI/CD Pipeline Into an Attack Surface", Jul 2026.
]-> Dataset and reproduction code

// Prompt secrecy and distributed verification are not security controls: a trusted agentic pipeline shipped attacker-specified secret exfiltration because an authority claim made verifiers rubber-stamp code whose malice they could see, while content controls were blind to it..

#tools
#MLSecOps
"The Trigger in the Haystack:
Extracting and Reconstructing LLM Backdoor Triggers", Feb. 2026.
]-> https://github.com/microsoft/llm-backdoor-scanner

#MLSecOps
#SCADA_Security
"Recording-Aware Evaluation of Multimodal OT Intrusion Detection", Oct 2026.
]-> https://github.com/ANTS-OntarioTechU/Recording-Aware-Evaluation-of-Multimodal-OT-Intrusion-Detection

#MLSecOps
"To Cache or Not to Cache: Understanding and Mitigating Safety Incompatibility in Cross-Model KV Cache Reuse", Oct 2026.

]-> VReparo - Value alignment for safer cross-model KV cache reuse