The AI Learning Hub
◆ Reference

AI Glossary

55 terms used across the courses, each defined in plain language and linked to the lessons that explain it.

Core AI vocabulary

Agent

AI system that pursues a goal with autonomy, takes actions via tools, and iterates based on outcomes — distinguished from a single-prompt…

Agentic AI

The broader shift from AI that answers questions to AI that does work — systems that plan, call tools, check their own results, and carry…

Attention

The mechanism at the heart of the transformer: for every token, the model weighs how relevant every other token in the context is, lettin…

Bias (in AI)

Systematic unfairness in model behavior, usually inherited from patterns in training data. Shows up as skewed outputs for certain groups,…

Context Window

Maximum number of tokens a model can process in one call (input + output). Modern frontier models offer 200K to 2M+ tokens.

Deepfake

Synthetic audio, image, or video that convincingly imitates a real person. Powered by the same generative models used for legitimate medi…

Distillation

Training a smaller, cheaper model to imitate a larger one's outputs. The main route by which frontier-level capability trickles down into…

Embedding

Numerical vector representation of content where semantically similar items cluster in high-dimensional space. Foundation of vector searc…

Eval (Evaluation)

A structured test suite for AI behavior — benchmarks, graded tasks, or automated judges that measure whether a model or agent actually do…

Fine-Tuning

Further training a base model on domain examples to alter its behavior. Compared to prompting: more expensive, slower to update, behavior…

Generative AI

AI that produces new content — text, images, audio, video, code — rather than only classifying or ranking existing content. LLMs and diff…

Guardrails

The layer of controls wrapped around a model: input screening, output filtering, blocked topics, structured formats, and approval gates. …

Hallucination

Fluent, confident model output not grounded in reality. Mitigated through grounding (RAG), citations, constrained outputs, and human-in-t…

HITL (Human-in-the-Loop)

Design pattern where humans approve or oversee AI actions, especially at consequential or irreversible decision points.

Inference

Running a trained model to get an answer — as opposed to training, which builds the model. Training happens rarely and costs a lot once; …

LLM (Large Language Model)

Neural network trained on massive text corpora to predict tokens. The substrate of modern generative AI.

MCP (Model Context Protocol)

Open standard (originated at Anthropic, broadly adopted) for AI hosts to discover and call tools, resources, and prompts via a uniform in…

Model Card

A short, structured document describing a model: intended use, training data summary, evaluation results, limitations, and known failure …

Multimodal

A model that works across more than one medium — reading images, hearing audio, or generating video as well as text. Frontier models are …

Open-Weight Model

A model whose trained weights are published for anyone to download, run, and fine-tune (Llama, DeepSeek, Qwen, Gemma families). Contraste…

Parameter

One learned number inside a neural network. Model "size" is the parameter count — billions to trillions of these values, tuned during tra…

Prompt

Everything you send a model in one request: the instruction, the context, the examples, the data. Prompt quality is the cheapest, fastest…

Prompt Injection

Malicious instructions embedded in inputs (often via retrieved data) that override intended model behavior. Indirect prompt injection is …

RAG (Retrieval-Augmented Generation)

Pattern of retrieving relevant content from a private knowledge base via vector search and feeding it into the LLM prompt at query time. …

Reasoning Model

A model trained (typically with reinforcement learning) to think before answering — generating internal chains of reasoning and spending …

System Prompt

The standing instructions an application gives the model before any user input — persona, rules, tools, and boundaries. Users rarely see …

Temperature

A sampling setting that controls how adventurous a model's word choices are. Low temperature: focused and repeatable. High temperature: v…

Token

Unit of input or output processed by an AI model. For text, typically 3-4 characters of English. Images are tokenized as patches, video a…

Tokenization

Chopping raw text (or images, or audio) into the discrete units a model actually processes. Explains many odd model behaviors — why count…

Training

The process of building a model: showing it enormous amounts of data and adjusting its parameters until its predictions improve. Followed…

Transformer

The neural network architecture behind modern AI, built on the attention mechanism. Introduced in 2017; nearly every frontier language mo…

Security & agent-platform terms

A2A (Agent2Agent Protocol)

Open protocol for agent-to-agent discovery and communication, championed by Google and now stewarded under the Linux Foundation. Companio…

ADK (Agent Development Kit)

Google's open-source framework for building AI agents in code. Agents built with it can run anywhere, including on Agent Engine, the mana…

Agent Gateway

The enforcement point for agent traffic in Google's Gemini Enterprise Agent Platform. It checks each agent-to-tool and agent-to-agent con…

Agent Registry

Google's central catalog of an organisation's agents and MCP servers in Gemini Enterprise Agent Platform. Agent Gateway uses it to decide…

AIBOM

AI Bill of Materials — inventory of models, training data, agents, and MCP connections in an AI system. Emerging compliance and security …

BEC (Business Email Compromise)

Email-based fraud where an attacker impersonates an executive, vendor, or trusted party to trick the target into wiring funds, sharing cr…

EDR (Endpoint Detection and Response)

Security tooling focused on the endpoint — process activity, file behavior, command-line forensics. Heavy users of classical ML for behav…

Gemini Enterprise Agent Platform

Google Cloud's platform for building, deploying, governing and optimising AI agents, announced at Google Cloud Next on 22 April 2026 as t…

Google SecOps

Google Security Operations: Google's cloud-native SIEM and SOAR platform, formerly Chronicle and Siemplify. In 2026 it added AI agents fo…

GTI (Google Threat Intelligence)

Unified threat intelligence offering combining Mandiant TI and VirusTotal capabilities.

ITDR (Identity Threat Detection and Response)

Security category focused on detecting attacks against identity systems — credential theft, session hijacking, MFA fatigue, lateral movem…

Mandiant

Frontline-incident-response firm acquired by Google; the source of Mandiant Consulting, Managed Defense, and the bulk of Google threat in…

MITRE ATLAS

Public knowledge base of adversary tactics and techniques targeting AI systems — the AI counterpart to MITRE ATT&CK. The neutral referenc…

Model Armor

Google Cloud's screening service for AI prompts and responses, detecting prompt injection, jailbreaks, sensitive-data leaks and harmful c…

NHI (Non-Human Identity)

Identities for service accounts, API keys, machines. Distinguished from agent identities, which are autonomous and goal-oriented.

OWASP Top 10 for LLM Applications

The industry-standard list of the most critical risks in LLM-powered applications — prompt injection, excessive agency, supply chain, dat…

SCC (Security Command Center)

Google Cloud's cloud security platform (CNAPP): posture management, threat detection and vulnerability management. Its AI Protection feat…

SIEM (Security Information and Event Management)

Centralized platform for collecting, normalizing, and analyzing security telemetry. Well-known examples include Splunk, Microsoft Sentine…

SOAR (Security Orchestration, Automation, and Response)

The playbook-and-automation layer that orchestrates response actions across security tools. Increasingly combined with AI agents that dec…

SPIFFE

Secure Production Identity Framework For Everyone — open standard for workload identity, increasingly used as the basis for giving AI age…

UDM (Unified Data Model)

Google SecOps normalized schema for security telemetry. Search, detection rules, and analytics operate over UDM.

UEBA (User and Entity Behavior Analytics)

Detection technique that builds behavioral baselines for users and machine entities, then flags deviations. Predominantly unsupervised ML…

XDR (Extended Detection and Response)

Security platform category that unifies detection across endpoint, network, identity, and cloud. The "extended" generally means EDR plus …

YARA-L 2.0

Detection rule language used by Google SecOps. Now supports multi-stage queries for sophisticated detection logic.