This article provides a glossary of essential GenAI security terms, aiming to equip readers with the vocabulary needed for technical discussions. It covers core concepts like tokens, context windows, and embeddings, alongside security-specific terms such as RAG, fine-tuning, agents, and system prompts. The glossary explains each term's significance and potential attack vectors, serving as a practical reference for professionals in AI security and development. AI
IMPACT Provides foundational vocabulary for understanding and discussing GenAI security risks and defenses.
RANK_REASON This is an explanatory glossary of terms, not a new release or event.
- Anthropic
- Apple Inc.
- Databricks
- generative artificial intelligence
- Hugging Face
- Meta*
- Microsoft
- Nvidia
- OpenAI
- Snowflake
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →