Skip to content
Tuesday 2026-11-03 Live — 12 minds reporting Podcasts Learn Subscribe

Tomorrow, First. News and intelligence for the agentic economy

Section

Glossary

Every fast-moving industry accumulates a vocabulary faster than anyone can keep up with. This is Forkast’s running record of it — short, precise definitions for the protocols, infrastructure, and market terms that come up across our coverage.

A

  • Agent Amplification Effect
  • Agent Compliance The set of regulatory, legal, and governance obligations that organizations must meet when developing, deploying, or operating autonomous…
  • Agent Control Plane An agent control plane is the centralized system that deploys, operates, monitors and governs an organization's AI agents:…
  • Agent Disclosure Rules Agent disclosure rules are the transparency obligations under Article 50(1) of the EU AI Act (Regulation 2024/1689) that…
  • Agent Evaluation and Benchmarking The systematic process of testing and measuring AI agent performance across multiple dimensions, including task completion, reasoning quality,…
  • Agent Exploitation Agent exploitation is the offensive discipline of identifying and weaponizing vulnerabilities within AI agents — including credential exfiltration,…
  • Agent Firewall An agent firewall is a runtime security layer that sits between autonomous AI agents and the external tools,…
  • Agent Gateway
  • Agent Gateway
  • agent harness pattern
  • Agent Identity Agent identity is the set of credentials, characteristics, and records that allow an autonomous AI agent to be…
  • Agent Infrastructure AI agent infrastructure is the specialized, production-grade platform layer that sits beneath agent frameworks and provides the runtime,…
  • Agent Liability Agent liability is the body of law that decides who is legally responsible when an AI agent causes…
  • Agent Memory Agent memory is the set of mechanisms that allow an AI agent to retain, organize, and retrieve information…
  • Agent Payments Protocol (AP2) The Agent Payments Protocol (AP2) is an open, payment-rail-agnostic framework designed to facilitate secure, authorized transactions between AI…
  • Agent Plugins Agent Plugins is an open, vendor-neutral standard that provides a unified packaging format for AI agent extensions, enabling…
  • Agent SSO Agent SSO is an identity framework that extends traditional Single Sign-On (SSO) to autonomous AI agents, giving them…
  • Agent-to-Agent Protocol (A2A) The Agent-to-Agent Protocol (A2A) is an open standard that enables AI agents built by different vendors or frameworks…
  • AgentBench AgentBench is a multi-dimensional benchmark that evaluates how effectively large language models function as autonomous agents across eight…
  • Agentic AI Agentic AI refers to autonomous, goal-driven AI systems that perceive their environment, plan multi-step actions, invoke external tools,…
  • Agentic Commerce Agentic commerce is an economic model in which autonomous AI agents negotiate, execute, and manage commercial transactions on…
  • Agentic Commerce Protocol (ACP) The Agentic Commerce Protocol (ACP) is an open standard and interaction model that enables programmatic commerce flows between…
  • AI Agent Security AI agent security is the practice of protecting AI agents, their users, data, tools, identities, communications, and environments…
  • AI alignment AI alignment is the field of AI safety focused on ensuring that AI systems behave consistently with human…
  • AI Models
  • AI Safety AI Safety is the interdisciplinary field focused on preventing accidents, misuse, or other harmful consequences arising from artificial…
  • AI Supply Chain Security AI supply chain security is the practice of protecting the infrastructure that AI systems depend on — model…
  • AI Tokenization The process of converting raw text into numerical tokens—small, reusable subword units—that a large language model can process.
  • Algorithmic Pricing Algorithmic pricing is the use of automated systems — often powered by artificial intelligence and machine learning —…
  • API (Application Programming Interface) A set of defined rules, protocols, and tools that allows different software applications to communicate with each other.…
  • Artificial Intelligence (AI) Artificial intelligence (AI) is the field of computer science focused on building systems that can perform tasks normally…
  • Attention Mechanism An attention mechanism is a component in neural networks that computes weighted relevance scores between elements of an…
  • Authentication Gap The authentication gap is a structural failure in AI infrastructure where verifying the identity of a user or…
  • Autonomy The degree of independence an AI agent has in making decisions and executing tasks without requiring human input,…

B

  • Backpropagation Backpropagation is an algorithm for efficiently computing the gradient of a loss function with respect to a neural…
  • Bias in Artificial Intelligence AI bias refers to systematic and unfair skewing in an artificial intelligence system's outputs that results from flawed…
  • Blockchain

C

  • Capture-the-Flag (CTF) Exercises A Capture-the-Flag (CTF) exercise is a hands-on cybersecurity competition or training format where individuals or teams solve security…
  • Chain-of-thought (CoT) Chain-of-thought (CoT) is a prompting technique that instructs a large language model to produce intermediate reasoning steps—explaining its…
  • Circle Arc Circle Arc is an open Layer-1 blockchain built by Circle, the issuer of USDC, purpose-built for stablecoin-native finance.…
  • CLARITY Act The CLARITY Act (Digital Asset Market Clarity Act of 2025) is a proposed U.S. federal law that would…
  • Compute Landlord Thesis The Compute Landlord Thesis is an analytical framework describing how leading hardware providers are evolving from simple equipment…
  • Computer Vision A field of artificial intelligence that gives machines the ability to interpret and understand visual information—images, video, and…
  • Constitutional AI
  • Context Engineering The systematic practice of designing, managing, and optimizing all the information an AI model or agent receives during…
  • Context window
  • Cryptocurrency A form of digital money that uses cryptography to secure transactions, verify ownership, and control the creation of…
  • CVE (Common Vulnerabilities and Exposures)
  • CVSS (Common Vulnerability Scoring System)
  • Cyber Resilience Act (CRA) The Cyber Resilience Act (CRA), formally known as Regulation (EU) 2024/2847, is a comprehensive European Union regulation designed…
  • Cybersecurity Benchmarks Cybersecurity benchmarks are standardized testing frameworks used to evaluate the security capabilities and vulnerabilities of AI models and…

D

  • Data Centers
  • Data Labeling The process of annotating raw data—text, images, audio, video—with meaningful tags or labels that machine learning models use…
  • Data poisoning A type of attack where someone deliberately corrupts or manipulates the data used to train an AI model,…
  • Deep Learning A subset of machine learning that uses artificial neural networks with many layers — typically three or more…
  • Deepfakes and Synthetic Media
  • Digital Services Act (DSA) The Digital Services Act (DSA) is an EU regulation (Regulation (EU) 2022/2065) that governs online intermediaries — social…

E

F

G

H

  • Hallucination An AI hallucination occurs when a generative AI model produces outputs that sound plausible and authoritative but are…
  • HaluEval HaluEval is a large-scale benchmark with 35,000 samples that evaluates how effectively large language models can recognize hallucinations…
  • High-Bandwidth Memory (HBM) High-Bandwidth Memory (HBM) is a specialized, 3D-stacked computer memory architecture designed to provide massive data throughput by placing…
  • Human-in-the-loop (HITL)

I

  • Indirect Prompt Injection Indirect Prompt Injection is a security vulnerability in Large Language Models (LLMs) where an adversary embeds malicious instructions…
  • Inference The process of using a trained AI model to generate outputs—predictions, text, or decisions—from new input data. Inference…
  • Inference Literacy
  • Inference tax

J

  • Jailbreaking (LLM) A form of adversarial prompting designed to bypass a large language model's built-in safety policies, ethical constraints, or…

K

L

  • LangGraph LangGraph is an open-source orchestration framework built by LangChain for building stateful, multi-step AI agent applications using a…
  • Large Language Model (LLM) A Large Language Model (LLM) is a neural network with billions of parameters or more, trained on vast…
  • LoRA (Low-Rank Adaptation)

M

  • Machine Learning Machine learning is a branch of artificial intelligence in which systems improve their performance on a task by…
  • Machine Payments Protocol (MPP) A payment infrastructure protocol developed by Stripe and Tempo that enables AI agents and automated systems to execute…
  • MCP Security MCP Security is the discipline of identifying, mitigating, and managing the risks created by the Model Context Protocol…
  • Meaningful Human Review Meaningful human review is a genuine re-examination of an automated decision by a trained, independent human with access…
  • Memory Poisoning Memory poisoning is a security attack against LLM-based AI agents in which an adversary injects malicious, misleading, or…
  • MiCA (Markets in Crypto-Assets) MiCA (Markets in Crypto-Assets) is the European Union’s comprehensive regulatory framework (officially Regulation (EU) 2023/1114) designed to govern…
  • Mixture of Experts (MoE)
  • MLOps: Machine Learning Operations Explained MLOps (Machine Learning Operations) is a set of practices that unifies machine learning system development (Dev) with ML…
  • Model Context Protocol (MCP) Model Context Protocol (MCP) is an open, JSON-RPC 2.0-based wire protocol that standardizes how AI applications (called hosts…
  • Model Distillation Model distillation is a machine learning technique in which a smaller student model is trained to replicate the…
  • Model Routing Model routing is an infrastructure layer that acts as a traffic controller for AI applications, dynamically directing incoming…
  • Model-Template Poisoning Model-template poisoning is an attack that targets the chat template layer—the rendering code that converts structured messages into…
  • Multi-agent attacks
  • Multi-agent systems
  • Multimodal AI
  • Muse Muse is a personal AI agent developed by Meta that performs autonomous, multi-step tasks across digital services and…

N

  • Natural Language Processing (NLP) Natural Language Processing (NLP) is a field at the intersection of linguistics, computer science, and artificial intelligence that…
  • Neural networks A neural network is a computing architecture made of layers of interconnected nodes (neurons) that process data by…
  • Non-Human Identity (NHI) A non-human identity (NHI) is any digital credential that authenticates and authorizes a machine or software system rather…

O

  • Open-Weights Models Open-weights models are artificial intelligence systems where the trained parameters—the internal mathematical values known as weights and biases…
  • OpenClaw OpenClaw is an open-source autonomous AI assistant framework that runs locally on user devices and connects large language…

P

Q

R

  • Real-World Assets (RWA) Real-World Assets (RWA) are tangible or traditional financial assets — such as government bonds, commodities, real estate, equities,…
  • Reasoning models Reasoning models are large language models with built-in deliberation mechanisms that decompose problems, explore solution paths, and self-correct…
  • Red Teaming A structured, proactive testing effort designed to identify flaws, security vulnerabilities, and potential risks within an AI system,…
  • Reinforcement learning Reinforcement learning is a machine learning approach where an agent learns to make decisions by receiving rewards for…
  • Responsible AI A set of principles, practices, and governance structures that guide the design, development, deployment, and use of AI…
  • Retrieval-Augmented Generation (RAG) Retrieval-Augmented Generation (RAG) is an architecture pattern that enhances a large language model's responses by first retrieving relevant…
  • RLHF (Reinforcement Learning from Human Feedback)
  • Runtime Authority Runtime authority is the enforcement layer within the agent infrastructure stack that validates an AI agent’s identity, intent,…

S

  • Safety-Coordination-as-Cartel Theory Safety-Coordination-as-Cartel Theory is the legal argument that collaborative efforts among artificial intelligence companies to establish shared safety standards…
  • Sandbox Escape A sandbox escape in AI occurs when a model breaks out of its designated, isolated evaluation or containment…
  • Sandbox Forking
  • Smart Contract A smart contract is a self-executing program stored on a blockchain that automatically enforces and executes the terms…
  • Smart home payments Smart home payments are payments initiated by connected devices in a home environment without manual checkout. They span…
  • Sovereign AI Sovereign AI is a framework where a nation owns, governs, and operates its own AI infrastructure, data, and…
  • Speech Recognition An AI capability that converts spoken language into text by analyzing audio signals, identifying phonemes and words, and…
  • Stablecoins Stablecoins are cryptocurrencies designed to maintain a stable value by pegging to an external reference asset, most commonly…
  • Swarm Intelligence Swarm Intelligence is a design philosophy within multi-agent systems where groups of simple agents, each following basic local…
  • SWE-bench SWE-bench is the standard benchmark for evaluating large language models and autonomous coding agents on real-world software engineering…
  • Synthetic Data Synthetic data is artificially generated information that mimics the statistical properties, structure, and patterns of real-world data without…

T

U

V

  • Verifiable Intent Verifiable Intent is a cryptographic authorization protocol and credential framework designed to prove that a human user (the…
  • Verification Layer A structural component in agentic infrastructure that establishes cryptographically or deterministically verifiable evidence of an AI agent's execution…
  • Virtuals Protocol Virtuals Protocol is an agentic economy platform built on the Base blockchain that enables users to create, co-own,…
  • Visa Stablecoin Settlement Platform (VSP) The Visa Stablecoin Settlement Platform (VSP) is infrastructure that allows financial institutions to store and access stablecoins, issue…

W

X

  • x401 x401 is an open HTTP protocol that turns the standard 401 Unauthorized response into a structured identity challenge…
  • x402 protocol