Machine learning

AI safety evolved: secure-by-design, safe-by-measurement

Introducing a new framework for evaluating autonomy in security AI

How Sublime's AI agents are secure by design

Sublime’s AI agents are just the tip of the platform

More than “plausible nonsense”: A rigorous eval for ADÉ, our security coding agent

Technical deep dive of NLU 3.0: Modular, multi-headed, with advanced synthetic training

How ASA thinks: The technical architecture of Sublime’s Autonomous Security Analyst

Email Topic Modeling: Simplifying detection with ML-powered granularity

Combating GenAI Email Attacks with BERT LLM

Sublime Attack Score: Explainable, AI-backed threat analysis

Unmasking BEC attacks using Natural Language Understanding + MQL