# Daily Bytes

> Two minutes a day on AI, security and the skills around them. Each byte is one short lesson; bytes are grouped into courses and subjects, with a plain-words AI glossary and practice exams. On the web and on phones, with no account. Made by Tecyfy (https://tecyfy.com).

Course and subject pages list their bytes. Reading the bytes themselves, quizzes and progress happen in the app.

## Courses

- [Foundations](https://dailybytes.tecyfy.com/course/foundations): Understand what the AI tools around you are doing, where they fail, and how to stay safe with them. No background needed.
- [How AI Works](https://dailybytes.tecyfy.com/course/how-ai-works): Know what a chatbot is really doing when it answers, why it can sound sure and still be wrong, and how to use it with that in mind.
- [Living and Working with AI](https://dailybytes.tecyfy.com/course/living-and-working-with-ai): Ask AI well, check what matters, and agree how your household uses it for health, money, school and errands.
- [Claude AI](https://dailybytes.tecyfy.com/course/claude-ai): Get more from Claude: its apps and models, memory and Projects, and how Anthropic recommends you prompt it.
- [Getting AI Right](https://dailybytes.tecyfy.com/course/getting-ai-right): Get answers from AI you can trust and use: brief it well, ask for what you need, and check what comes back.
- [AI-Powered Phishing](https://dailybytes.tecyfy.com/course/ai-powered-phishing): You don't need to be good at spotting fakes. Learn what every scam needs from you, lock your logins so they can't be phished, and know what to do in the first hour if something goes wrong.
- [Security in the Age of AI](https://dailybytes.tecyfy.com/course/security-in-the-age-of-ai): Know what AI changed for attackers, the threats inside your own AI features, how to handle an incident, and how to write AI policy people will follow.
- [Code and Build with AI](https://dailybytes.tecyfy.com/course/building-with-ai-agents): Hand real work to a coding agent without losing control of your code, then put models into real software: cost, tool use, retrieval, evals, and agents you can trust in production.
- [Prepare for the Claude Certified Architect – Foundations exam](https://dailybytes.tecyfy.com/course/claude-architect-foundations-prep): Walk into the CCAR-F exam knowing how it marks answers: the 5 domains, the guide's 4 build exercises and exam-style practice inside its scenarios. Daily Bytes is independent of Anthropic.

## Subjects

- [AI Agents & Automation](https://dailybytes.tecyfy.com/topic/ai-agents-and-automation): What changes when an AI stops answering and starts doing: tools, permissions, limits, and the checks to set before you let one act for you.
- [AI Assistants in Daily Life](https://dailybytes.tecyfy.com/topic/ai-assistants-in-daily-life): Getting real help from AI at home: asking well, checking what matters, and agreeing how your household uses it for health, money, school and errands.
- [AI Policy & Data Protection](https://dailybytes.tecyfy.com/topic/ai-policy-and-data-protection): Shadow AI, data leaving through the prompt box, vendor promises and a policy people will follow: managing AI risk in an organisation without fear.
- [AI-Powered Phishing & Social Engineering](https://dailybytes.tecyfy.com/topic/ai-powered-phishing): Phishing after the typos disappeared: judging the request instead of the writing, verifying out of band, and building a culture that reports fast.
- [Building with LLM APIs](https://dailybytes.tecyfy.com/topic/building-with-llm-apis): For developers putting models into real software: tokens and cost, structured output, tool use, retrieval, evals, and agents you can trust in production.
- [Claude AI](https://dailybytes.tecyfy.com/topic/claude-ai): Claude itself, as of October 2026: its apps and models, memory and Projects, and how Anthropic recommends you prompt it.
- [Claude Architect Exam Prep](https://dailybytes.tecyfy.com/topic/claude-architect-foundations): Preparation for the Claude Certified Architect – Foundations exam: every task statement, the 4 build exercises and exam-style practice. Daily Bytes is independent of Anthropic.
- [Coding with AI Agents](https://dailybytes.tecyfy.com/topic/coding-with-ai-agents): Working with a coding agent without losing control of your codebase: tests as the spec, small changes, honest review, and knowing when to take the keyboard back.
- [Data Literacy & Charts](https://dailybytes.tecyfy.com/topic/data-literacy-and-charts): Reading charts and claims honestly, whoever made them: axes, baselines, missing data, and the jump from 'linked to' to 'causes'.
- [Deepfakes & Misinformation](https://dailybytes.tecyfy.com/topic/deepfakes-and-misinformation): Telling real from generated, and true from merely convincing: what you can actually check about a video, a photo or a story.
- [Deepfakes & Voice Clones](https://dailybytes.tecyfy.com/topic/deepfakes-and-voice-clones): Cloned voices and faked video calls aimed at payments and access, and the verification rules that hold up under pressure.
- [Getting AI Right](https://dailybytes.tecyfy.com/topic/getting-ai-right): Getting answers from AI you can trust and use: briefing it well, asking for what you need, and checking what comes back.
- [How AI Works](https://dailybytes.tecyfy.com/topic/how-ai-works): What a chatbot is really doing when it answers, why it can sound sure and be wrong, and what that means for how you use it.
- [Identity, Passkeys & MFA](https://dailybytes.tecyfy.com/topic/identity-passkeys-and-mfa): Which sign-in methods a phishing kit can defeat and which it can't, and how to move people and organisations to the ones it can't.
- [Incident Basics](https://dailybytes.tecyfy.com/topic/incident-basics): Investigating and responding when AI is part of the story: using it in the SOC without trusting it blindly, the logs you'll need, and incidents an assistant causes.
- [Leading Teams Through AI](https://dailybytes.tecyfy.com/topic/leading-teams-through-ai): Bringing AI into a team: who checks what, who decides, and how one person's shortcut becomes a reliable shared way of working.
- [Online Scams in the AI Era](https://dailybytes.tecyfy.com/topic/online-scams): The scams AI made cheaper and more convincing, from cloned voices to fake helplines, and the few habits that beat all of them.
- [Privacy & Your Data](https://dailybytes.tecyfy.com/topic/privacy-and-your-data): Getting help from AI without handing over what should stay private: what never to paste, how to strip a file down, and what privacy settings really do.
- [Secure Coding with AI Assistants](https://dailybytes.tecyfy.com/topic/secure-coding-with-ai): Building AI features that survive contact with attackers: prompt injection, untrusted output, secrets, and the dependencies a model invents.
- [Securing AI Agents & Tools](https://dailybytes.tecyfy.com/topic/securing-ai-agents-and-tools): Agents as a new attack surface: what they can touch, whose permissions they use, poisoned tools and memory, and the approvals and monitoring that keep them in bounds.
- [Threat Modelling](https://dailybytes.tecyfy.com/topic/threat-modelling): Thinking like an attacker now that attackers have AI: what it changed and what it left alone, the OWASP risks in plain terms, why injection can't simply be patched, how to test your own feature, and which flaws to fix first.

## Practice exams

- [Claude Architect – Foundations mock exam](https://dailybytes.tecyfy.com/exam/claude-architect-foundations-mock): 60 questions, 120 minutes, with Plus

## AI glossary

- [AI glossary](https://dailybytes.tecyfy.com/glossary): plain meanings for the words you meet when you use AI
- [Access review](https://dailybytes.tecyfy.com/glossary/access-review): A check in which the owner of a site, group or system confirms who still needs access and removes everyone who doesn't.
- [Account takeover](https://dailybytes.tecyfy.com/glossary/account-takeover): When someone else gets into one of your accounts and uses it as you, to steal money, read your messages or scam your contacts.
- [Agent identity](https://dailybytes.tecyfy.com/glossary/agent-identity): An account made for one AI agent, separate from any person's login, so its actions can be traced and it can be switched off alone.
- [Agent instructions file](https://dailybytes.tecyfy.com/glossary/agent-instructions-file): A file in your repository that a coding agent reads every session: the commands, conventions and gotchas it can't work out from the code.
- [Agent trace](https://dailybytes.tecyfy.com/glossary/agent-trace): A step-by-step record of one AI agent run: every model call, every tool call and its result, under one run ID.
- [Agentic loop](https://dailybytes.tecyfy.com/glossary/agentic-loop): The cycle behind an AI agent: it asks for a tool, your code runs it, the result goes back, and it decides the next step.
- [AGI](https://dailybytes.tecyfy.com/glossary/agi): Artificial general intelligence: AI that could match or beat people across most kinds of work. A stated goal of several labs, with no agreed definition.
- [AI agent](https://dailybytes.tecyfy.com/glossary/ai-agent): An AI that carries out a task in steps, using tools such as a browser, your files or apps, instead of only replying.
- [AI bias](https://dailybytes.tecyfy.com/glossary/ai-bias): When an AI system's outputs systematically favour or disadvantage some people or views, usually because of patterns in its training data.
- [AI browser](https://dailybytes.tecyfy.com/glossary/ai-browser): A web browser with a built-in AI assistant that reads your pages and, in agent mode, clicks and fills in forms for you.
- [AI companion](https://dailybytes.tecyfy.com/glossary/ai-companion): A chatbot built to be a friend, partner or confidant, with a persistent character that remembers you and keeps the conversation going.
- [AI detector](https://dailybytes.tecyfy.com/glossary/ai-detector): A tool that estimates whether text, images or audio were made by AI. For text, its results are unreliable.
- [AI gateway](https://dailybytes.tecyfy.com/glossary/ai-gateway): Software that sits between an organisation's apps and the AI models, holding the provider keys and routing, logging and limiting every request.
- [AI in group chats](https://dailybytes.tecyfy.com/glossary/group-chat): An AI assistant you can bring into a chat with other people, where it answers questions for everyone to see.
- [AI Mode](https://dailybytes.tecyfy.com/glossary/ai-mode): Google Search's chat-style mode: you ask longer questions, get a written answer with links and ask follow-ups, like a chatbot that searches.
- [AI model](https://dailybytes.tecyfy.com/glossary/ai-model): The trained system inside an AI app that does the actual work: a large file of learned numbers that turns your input into an output.
- [AI notetaker](https://dailybytes.tecyfy.com/glossary/ai-notetaker): An AI tool that records or transcribes a meeting and writes the summary and action points, often by joining the call as an extra participant.
- [AI Overviews](https://dailybytes.tecyfy.com/glossary/ai-overviews): The AI-written answer Google shows at the top of many search results, built from web pages it links to.
- [AI slop](https://dailybytes.tecyfy.com/glossary/ai-slop): Low-quality content made in bulk with AI: filler articles, odd images, junk videos and padded reports that look like work but say little.
- [AI watermark](https://dailybytes.tecyfy.com/glossary/ai-watermark): A hidden signal built into AI-generated images, video, audio or text so software can later tell it was made by AI.
- [Alert triage](https://dailybytes.tecyfy.com/glossary/alert-triage): The first sort of security alerts: which are real, which are noise, and which need someone now.
- [Alignment](https://dailybytes.tecyfy.com/glossary/alignment): Making an AI system reliably do what its makers and users actually intend, including refusing what it shouldn't do.
- [Answer engine](https://dailybytes.tecyfy.com/glossary/answer-engine): A search tool that reads web pages for you and writes one answer with citations, instead of giving you a list of links.
- [API](https://dailybytes.tecyfy.com/glossary/api): The way one program talks to another. AI companies sell model access through APIs, so other apps can build their features on top.
- [API key](https://dailybytes.tecyfy.com/glossary/api-key): A secret code that lets a program use an AI service on your account. Anyone who has it can run up charges in your name.
- [APP fraud](https://dailybytes.tecyfy.com/glossary/app-fraud): Fraud where you are tricked into sending a bank transfer yourself, to an account you think is genuine.
- [Artificial intelligence](https://dailybytes.tecyfy.com/glossary/artificial-intelligence): A broad label for computer systems that do tasks which used to need human judgement, from filtering spam to writing text.
- [Attack surface](https://dailybytes.tecyfy.com/glossary/attack-surface): Every point where an outsider can try to get into a system or get data out of it: login pages, open ports, APIs, test servers.
- [Attention](https://dailybytes.tecyfy.com/glossary/attention): The step inside a transformer that decides which other words in the text matter most for understanding each word.
- [Audit log](https://dailybytes.tecyfy.com/glossary/audit-log): A time-ordered record of who did what on a system, kept so that actions can be checked and reconstructed later.
- [Autocomplete](https://dailybytes.tecyfy.com/glossary/autocomplete): AI that suggests your next words as you type in email, messages, search boxes or code; a tap or the Tab key accepts them.
- [Batch processing](https://dailybytes.tecyfy.com/glossary/batch-processing): Sending many model requests as one background job, at half the usual price, with results back within 24 hours.
- [Benchmark](https://dailybytes.tecyfy.com/glossary/benchmark): A fixed test used to score and compare AI models, like an exam where every model gets the same questions.
- [Breach notification](https://dailybytes.tecyfy.com/glossary/breach-notification): A legal duty to tell a regulator, and sometimes the people affected, about a security incident within a set time.
- [Bug bounty](https://dailybytes.tecyfy.com/glossary/bug-bounty): A standing offer to pay outside researchers who find and privately report security flaws in an organisation's software.
- [Business email compromise](https://dailybytes.tecyfy.com/glossary/business-email-compromise): A scam where criminals pose as a boss, colleague or supplier, usually by email, to get a business to pay them or change bank details.
- [Call-back check](https://dailybytes.tecyfy.com/glossary/callback-verification): Checking an unexpected request by contacting the person or organisation yourself, on a number or channel you already trust.
- [Canvas](https://dailybytes.tecyfy.com/glossary/canvas): A side panel where you and the AI edit one document or piece of code together, instead of getting a fresh copy in every reply.
- [Chain of thought](https://dailybytes.tecyfy.com/glossary/chain-of-thought): The step-by-step working a model writes before its final answer, either because a prompt asked for it or because it was trained to.
- [Chat history](https://dailybytes.tecyfy.com/glossary/chat-history): The record of your past conversations that an AI app keeps, so you can find, continue or delete them.
- [Chatbot](https://dailybytes.tecyfy.com/glossary/chatbot): An app you talk to in plain language, which answers using an AI model running behind it.
- [Citation](https://dailybytes.tecyfy.com/glossary/citation): A link or reference an AI attaches to a claim to show where it came from. Worth checking, not proof on its own.
- [ClickFix](https://dailybytes.tecyfy.com/glossary/clickfix): A trick where a web page or video tells you to copy a command and paste it into your computer, which then installs malware.
- [Coding agent](https://dailybytes.tecyfy.com/glossary/coding-agent): An AI that works on a software project by itself: it reads the code, makes changes across files, runs tests and reports back.
- [Compaction](https://dailybytes.tecyfy.com/glossary/compaction): When a coding agent replaces the earlier part of a long session with a summary, to make room in its context window.
- [Computer use](https://dailybytes.tecyfy.com/glossary/computer-use): An AI that operates a computer the way you do: it looks at the screen, moves the pointer, clicks and types to finish a task.
- [Confidence calibration](https://dailybytes.tecyfy.com/glossary/confidence-calibration): How closely a model's stated confidence matches how often it's actually right, and the work of setting review thresholds so that it does.
- [Confused deputy](https://dailybytes.tecyfy.com/glossary/confused-deputy): A program with more access than the person using it, tricked into using that access for them.
- [Connector](https://dailybytes.tecyfy.com/glossary/connector): A link that lets an AI app read from, and sometimes act in, another service you use, like your email, calendar or files.
- [Containment](https://dailybytes.tecyfy.com/glossary/containment): The first step of an incident response: stopping the harm from spreading before anyone starts on the cause.
- [Content Credentials](https://dailybytes.tecyfy.com/glossary/content-credentials): A record attached to an image, video or audio file showing how it was made and edited, including whether AI was used.
- [Context engineering](https://dailybytes.tecyfy.com/glossary/context-engineering): Choosing everything an AI sees for a task (instructions, files, past messages, tool results), not just the wording of the request.
- [Context rot](https://dailybytes.tecyfy.com/glossary/context-rot): The way a model's answers get worse as its input grows longer, often well before the context window is full.
- [Context window](https://dailybytes.tecyfy.com/glossary/context-window): How much text a model can take into account at once: the chat so far, files and instructions together.
- [Coordinator agent](https://dailybytes.tecyfy.com/glossary/coordinator-agent): In a multi-agent system, the agent that splits a job into parts, hands each to a sub-agent and combines what comes back.
- [Copilot](https://dailybytes.tecyfy.com/glossary/copilot): Microsoft's name for its AI assistants, and a general word for an AI that works alongside you while you stay in charge.
- [Credential harvesting](https://dailybytes.tecyfy.com/glossary/credential-harvesting): Collecting usernames, passwords, tokens and keys in bulk, to break in with them later or to sell them.
- [Credential stuffing](https://dailybytes.tecyfy.com/glossary/credential-stuffing): Trying usernames and passwords leaked from one site on many other sites, automatically, to break into accounts where people reused them.
- [Custom GPT](https://dailybytes.tecyfy.com/glossary/custom-gpt): A saved, named version of ChatGPT with its own instructions and files. Custom GPTs are being retired in favour of plugins.
- [Custom instructions](https://dailybytes.tecyfy.com/glossary/custom-instructions): Standing instructions you write once, which the app adds to every new chat: who you are and how you want answers.
- [CVE](https://dailybytes.tecyfy.com/glossary/cve): The public reference number given to a known security flaw, such as CVE-2025-32711, so that vendors, scanners and defenders all mean the same flaw.
- [CVSS](https://dailybytes.tecyfy.com/glossary/cvss): A score from 0 to 10 for how severe a security flaw is. It describes the flaw, not whether anyone is attacking it.
- [Data breach](https://dailybytes.tecyfy.com/glossary/data-breach): When information a company holds about people, such as emails, passwords, addresses or chats, gets exposed or stolen.
- [Data loss prevention](https://dailybytes.tecyfy.com/glossary/data-loss-prevention): Software that spots sensitive data being copied, uploaded or sent somewhere it shouldn't go, and warns the person or stops it.
- [Data poisoning](https://dailybytes.tecyfy.com/glossary/data-poisoning): Planting doctored material in what an AI learns from, so the model later misbehaves in a way the attacker chose.
- [Deep learning](https://dailybytes.tecyfy.com/glossary/deep-learning): Machine learning with neural networks that have many layers: the approach behind today's chatbots, image generators and voice assistants.
- [Deep research](https://dailybytes.tecyfy.com/glossary/deep-research): A mode where the AI spends minutes running many searches and writes a long report with citations.
- [Deepfake](https://dailybytes.tecyfy.com/glossary/deepfake): AI-made or AI-altered video, image or audio showing a real person saying or doing something they never did.
- [Diffusion model](https://dailybytes.tecyfy.com/glossary/diffusion-model): A generator that starts from random noise and removes it step by step until an image, video or sound matches the prompt.
- [Disinformation](https://dailybytes.tecyfy.com/glossary/disinformation): False information spread on purpose to mislead people; misinformation is false information shared by people who believe it.
- [Distillation](https://dailybytes.tecyfy.com/glossary/distillation): Training a small model to copy a big model's answers, so it keeps much of the skill at a fraction of the size and cost.
- [Doxxing](https://dailybytes.tecyfy.com/glossary/doxxing): Finding and publishing someone's private details, such as home address, workplace or phone number, to harass or frighten them.
- [Edge device](https://dailybytes.tecyfy.com/glossary/edge-device): Equipment that sits between an organisation's network and the internet, such as a firewall, VPN gateway or router, and can be reached from outside.
- [eDiscovery](https://dailybytes.tecyfy.com/glossary/ediscovery): Finding, preserving and handing over electronic records for a legal case, an audit or an investigation, and the tools used to do it.
- [Embedding](https://dailybytes.tecyfy.com/glossary/embedding): A list of numbers standing for the meaning of some text or an image, so a computer can measure how alike 2 things are.
- [Escalation (human handoff)](https://dailybytes.tecyfy.com/glossary/escalation): When an AI agent passes a case to a person, along with what that person needs to act on it.
- [EU AI Act](https://dailybytes.tecyfy.com/glossary/eu-ai-act): The European Union's law on AI systems, Regulation (EU) 2024/1689, whose strictest rules fall on systems it classes as high-risk.
- [Eval](https://dailybytes.tecyfy.com/glossary/eval): A repeatable test for an AI feature: fixed cases, a pass/fail check on each, and a score you track every time something changes.
- [Fake AI app](https://dailybytes.tecyfy.com/glossary/fake-ai-app): A website, app or download dressed up as a popular AI tool, built to steal your logins, money or data.
- [Fake reviews](https://dailybytes.tecyfy.com/glossary/fake-reviews): Reviews written by sellers, paid writers, bots or AI to make a product look better, or a rival look worse, than it is.
- [False positive](https://dailybytes.tecyfy.com/glossary/false-positive): A flag raised on something that's actually fine: a review comment on correct code, or an alert on a harmless email.
- [Family emergency scam](https://dailybytes.tecyfy.com/glossary/family-emergency-scam): A call or message that seems to come from a relative in trouble, asking for money urgently and often for secrecy.
- [Few-shot prompting](https://dailybytes.tecyfy.com/glossary/few-shot-prompting): Showing an AI a few examples of what you want, so it copies the pattern instead of guessing it.
- [File upload](https://dailybytes.tecyfy.com/glossary/file-upload): Giving an AI a document, spreadsheet, photo or recording to work from, so it answers from your material rather than from memory.
- [Fine-tuning](https://dailybytes.tecyfy.com/glossary/fine-tuning): Extra training that adjusts an existing model for a style, format or task, using a smaller set of examples.
- [Foundation model](https://dailybytes.tecyfy.com/glossary/foundation-model): A large model trained on broad data once, then adapted for many different products instead of each product training its own.
- [Frontier model](https://dailybytes.tecyfy.com/glossary/frontier-model): One of the most capable AI models of its moment, from the labs pushing hardest, and the kind most safety laws are written for.
- [Generative AI](https://dailybytes.tecyfy.com/glossary/generative-ai): AI that makes new text, images, audio or video from patterns it learned, rather than only sorting or labelling things.
- [Git worktree](https://dailybytes.tecyfy.com/glossary/git-worktree): A second working folder for the same git repository, on its own branch, so 2 agents or tasks never edit the same files.
- [GPU](https://dailybytes.tecyfy.com/glossary/gpu): The chip that does AI's heavy maths. Built for video games, it now fills data centres and is the main cost of running AI.
- [Grounding](https://dailybytes.tecyfy.com/glossary/grounding): Tying a model's answer to specific sources, such as web pages or your own files, instead of what it remembers from training.
- [Guardrails](https://dailybytes.tecyfy.com/glossary/guardrails): The rules and filters that limit what an AI will say or do, from trained-in refusals to checks on its answers.
- [Hallucination](https://dailybytes.tecyfy.com/glossary/hallucination): When an AI states something false or made up as if it were fact, often fluently and with confidence.
- [Headless mode](https://dailybytes.tecyfy.com/glossary/headless-mode): Running a coding agent from a script or pipeline with nobody at the keyboard: it takes a prompt, works, prints the result and exits.
- [Hooks (agent hooks)](https://dailybytes.tecyfy.com/glossary/agent-hooks): Scripts a coding agent runs automatically at set points, such as before a tool runs or after a file is edited, every time.
- [Human in the loop](https://dailybytes.tecyfy.com/glossary/human-in-the-loop): A design in which an AI agent's action waits for a person's approval before it happens.
- [Hybrid search](https://dailybytes.tecyfy.com/glossary/hybrid-search): Search that runs keyword matching and embedding search on the same question and merges the results, so exact codes and loose wording both match.
- [Identity proofing](https://dailybytes.tecyfy.com/glossary/identity-proofing): Confirming a person is who they claim to be before giving them an account or access, usually by matching a face to an ID document.
- [Image generation](https://dailybytes.tecyfy.com/glossary/image-generation): Making a new picture from a written description, or changing a photo by describing the edit, using an AI model.
- [Impersonation scam](https://dailybytes.tecyfy.com/glossary/impersonation-scam): A scam where the caller or sender pretends to be someone you trust: your bank, the police, a tax office or a well-known company.
- [Independent review](https://dailybytes.tecyfy.com/glossary/independent-review): Having a separate AI instance, one that didn't write the work and doesn't share its reasoning, review it.
- [Indicator of compromise](https://dailybytes.tecyfy.com/glossary/indicator-of-compromise): A piece of technical evidence that a system may have been attacked: an address, a domain, a file's fingerprint, an account.
- [Indirect prompt injection](https://dailybytes.tecyfy.com/glossary/indirect-prompt-injection): Prompt injection that arrives in something the AI reads while working, such as an email, web page or file, not from its user.
- [Inference](https://dailybytes.tecyfy.com/glossary/inference): Running a trained model to get an answer. Training builds the model once; inference happens every time anyone asks it something.
- [Infostealer](https://dailybytes.tecyfy.com/glossary/infostealer): Malware that quietly copies saved passwords, browser cookies, card details and crypto wallets from a computer and sends them to criminals.
- [Initial access](https://dailybytes.tecyfy.com/glossary/initial-access): The first step of a break-in: how an attacker gets from outside an organisation to a first foothold inside it.
- [Injection attack](https://dailybytes.tecyfy.com/glossary/injection-attack): Feeding forged video or audio straight into an identity check or call, bypassing the camera or microphone. Not the same thing as prompt injection.
- [Interpretability](https://dailybytes.tecyfy.com/glossary/interpretability): Research into what is actually happening inside a model: which internal features and pathways produce a given answer.
- [Investment scam](https://dailybytes.tecyfy.com/glossary/investment-scam): A fake opportunity to grow your money, often crypto, sold through ads, celebrity endorsements or a trusted contact, where the money simply disappears.
- [Jailbreak](https://dailybytes.tecyfy.com/glossary/jailbreak): A trick to get an AI to ignore its safety rules and produce something it would normally refuse.
- [JSON schema](https://dailybytes.tecyfy.com/glossary/json-schema): A written contract for a JSON record: which fields it has, the type of each, which are required, and which values are allowed.
- [KEV catalog](https://dailybytes.tecyfy.com/glossary/kev-catalog): A public list, kept by the US cybersecurity agency, of security flaws attackers are known to have used: the ones to patch first.
- [Kill switch](https://dailybytes.tecyfy.com/glossary/kill-switch): A way to stop an AI feature or agent quickly when it goes wrong, set up and tried before it is needed.
- [Knowledge cutoff](https://dailybytes.tecyfy.com/glossary/knowledge-cutoff): The date a model's training data ends. It knows little or nothing about what happened after that.
- [Large language model](https://dailybytes.tecyfy.com/glossary/large-language-model): The kind of AI model behind chatbots: trained on vast amounts of text to predict what comes next in a sentence.
- [Least privilege](https://dailybytes.tecyfy.com/glossary/least-privilege): Giving a person, program or AI agent only the access its task needs, so a mistake or a takeover can do less damage.
- [Legal hold](https://dailybytes.tecyfy.com/glossary/legal-hold): An instruction to keep records that may matter to a dispute or investigation, which overrides normal deletion for as long as it stands.
- [Lethal trifecta](https://dailybytes.tecyfy.com/glossary/lethal-trifecta): An AI agent that has your private data, reads content from strangers, and has a way to send data out can be tricked into leaking.
- [Live translation](https://dailybytes.tecyfy.com/glossary/live-translation): AI that translates a spoken conversation as it happens, so 2 people without a shared language can talk.
- [Liveness check](https://dailybytes.tecyfy.com/glossary/liveness-check): A check that the face in a selfie or video belongs to a real person present right now, not a photo, mask or deepfake.
- [LLM-as-judge](https://dailybytes.tecyfy.com/glossary/llm-as-judge): Using a language model to grade another model's answers, for checks that take judgement rather than a rule code can apply.
- [LLMjacking](https://dailybytes.tecyfy.com/glossary/llmjacking): Using stolen AI keys or a hijacked cloud account to run someone else's AI work, on the victim's bill and under their name.
- [Logprobs](https://dailybytes.tecyfy.com/glossary/logprobs): Numbers showing how likely a model rated each token it wrote, and the runners-up, sent back to your code when you ask.
- [Lookalike domain](https://dailybytes.tecyfy.com/glossary/lookalike-domain): A web address made to look like a real one, with a swapped letter or extra word, used for fake login and shopping pages.
- [LoRA](https://dailybytes.tecyfy.com/glossary/lora): A cheaper way to fine-tune a model: its own weights stay frozen, and a small add-on called an adapter is trained beside them.
- [Lost in the middle](https://dailybytes.tecyfy.com/glossary/lost-in-the-middle): A model's tendency to use what sits at the start and end of a long input reliably, and to miss findings in the middle.
- [Machine learning](https://dailybytes.tecyfy.com/glossary/machine-learning): Teaching a computer by showing it examples, so it learns the pattern itself instead of following rules written by hand.
- [Malvertising](https://dailybytes.tecyfy.com/glossary/malvertising): Online ads that lead to scams or malware, often bought on real platforms and dressed up as trusted brands.
- [MCP resource](https://dailybytes.tecyfy.com/glossary/mcp-resource): Data an MCP server offers at an address, such as a database schema or open issues, for an app to read rather than call.
- [Memory](https://dailybytes.tecyfy.com/glossary/ai-memory): A feature that lets a chatbot keep notes about you, or draw on past chats, and use them in new conversations.
- [Memory poisoning](https://dailybytes.tecyfy.com/glossary/memory-poisoning): Planting false information in what an AI agent remembers, so that it keeps shaping the agent's decisions in later sessions.
- [Mixture of experts](https://dailybytes.tecyfy.com/glossary/mixture-of-experts): A model split into many sub-networks, only a few of which switch on for each word, so it can be huge yet cheap to run.
- [Model card](https://dailybytes.tecyfy.com/glossary/model-card): A document a lab publishes with a model, describing what it's for, how it was tested, its known limits and its risks.
- [Model collapse](https://dailybytes.tecyfy.com/glossary/model-collapse): Decline in a model's quality when it's trained, generation after generation, on AI-made data in place of human-made data.
- [Model Context Protocol](https://dailybytes.tecyfy.com/glossary/model-context-protocol): An open standard for plugging AI apps into other tools and data, so one connector can work with many assistants.
- [Model retirement](https://dailybytes.tecyfy.com/glossary/model-retirement): When an AI company removes an older model from its app or API, so nobody can choose it and chats move to a newer one.
- [Money mule](https://dailybytes.tecyfy.com/glossary/money-mule): Someone whose bank account is used to move stolen money, often recruited through a fake job or a "quick cash" offer.
- [Multimodal](https://dailybytes.tecyfy.com/glossary/multimodal): Able to take in or produce more than text: images, audio, video or files.
- [N-day](https://dailybytes.tecyfy.com/glossary/n-day): A security flaw attacked after its fix is public. The "n" is the number of days the fix has been available and not applied.
- [Neural network](https://dailybytes.tecyfy.com/glossary/neural-network): A kind of machine-learning model made of layers of simple connected units, whose connection strengths are tuned during training.
- [Next-word prediction](https://dailybytes.tecyfy.com/glossary/next-word-prediction): How a language model writes: it keeps picking a likely next piece of text, given everything before it.
- [Non-human identity](https://dailybytes.tecyfy.com/glossary/non-human-identity): Any account or credential that software uses in place of a person, such as a service account, an access key or an AI agent's identity.
- [Nudify app](https://dailybytes.tecyfy.com/glossary/nudify-app): An app or website that uses AI to make a fake nude image of a real person from an ordinary photo, without their consent.
- [Open-weight model](https://dailybytes.tecyfy.com/glossary/open-weight-model): A model whose trained weights anyone can download and run on their own computer or servers.
- [Oversharing](https://dailybytes.tecyfy.com/glossary/oversharing): When more people can open a file, folder or site than need to, through wide groups, company-wide links or access nobody removed.
- [OWASP Top 10](https://dailybytes.tecyfy.com/glossary/owasp-top-10): A widely used list of the 10 most critical security risks for web applications, with sister lists for AI applications and for AI agents.
- [Parameters](https://dailybytes.tecyfy.com/glossary/parameters): The billions of internal numbers a model adjusts during training; together they hold what it learned.
- [Passkey](https://dailybytes.tecyfy.com/glossary/passkey): A sign-in that uses your device's screen lock (face, fingerprint or PIN) instead of a password, and only works on the real site.
- [Password manager](https://dailybytes.tecyfy.com/glossary/password-manager): An app that creates, stores and fills a different strong password for every site, and won't fill them on a fake one.
- [Path-scoped rules](https://dailybytes.tecyfy.com/glossary/path-scoped-rules): Instruction files in Claude Code's .claude/rules/ folder that load only when Claude works on files matching their patterns.
- [Phishing](https://dailybytes.tecyfy.com/glossary/phishing): A message or call pretending to be someone you trust, to get you to click, pay, log in or share a code.
- [Phishing kit](https://dailybytes.tecyfy.com/glossary/phishing-kit): Ready-made software sold to criminals that sets up fake login or payment pages and collects what victims type, no skill needed.
- [Pig butchering](https://dailybytes.tecyfy.com/glossary/pig-butchering): A long investment scam, often run from industrial scam centres, where a new online friend slowly persuades you into a fake crypto platform.
- [Plan mode](https://dailybytes.tecyfy.com/glossary/plan-mode): A coding agent setting where it reads the code and writes out what it means to change, for you to approve before it edits.
- [Plugin](https://dailybytes.tecyfy.com/glossary/plugin): An add-on package for an AI app that bundles reusable instructions, reference files and connections to other services, installed from a directory.
- [Post-training](https://dailybytes.tecyfy.com/glossary/post-training): Everything done after pretraining to turn a raw text predictor into a helpful assistant: example conversations, ratings, rules and safety training.
- [Presentation attack](https://dailybytes.tecyfy.com/glossary/presentation-attack): Fooling a face, voice or fingerprint check by showing its sensor a fake: a photo, a mask, or a video playing on a screen.
- [Pretraining](https://dailybytes.tecyfy.com/glossary/pretraining): The first and biggest stage of training, where a model learns to predict the next word across an enormous amount of text.
- [Projects](https://dailybytes.tecyfy.com/glossary/projects): A workspace in an AI app that keeps related chats, files and instructions together, so every chat in it starts with that background.
- [Prompt](https://dailybytes.tecyfy.com/glossary/prompt): What you type or say to an AI in a turn: your question, your instructions and anything you attach.
- [Prompt caching](https://dailybytes.tecyfy.com/glossary/prompt-caching): The provider keeping a recent copy of a request's opening text, so sending the same text again costs a fraction of reading it fresh.
- [Prompt chaining](https://dailybytes.tecyfy.com/glossary/prompt-chaining): Splitting a task into a fixed series of model calls, each with its own prompt, where one step's output feeds the next.
- [Prompt engineering](https://dailybytes.tecyfy.com/glossary/prompt-engineering): The skill of writing requests that get an AI to do what you actually want: the context, the task, the format and the reasons.
- [Prompt injection](https://dailybytes.tecyfy.com/glossary/prompt-injection): Instructions hidden in something an AI reads, like a web page, email or file, that try to make it do something else.
- [Proof of concept](https://dailybytes.tecyfy.com/glossary/proof-of-concept): A small demonstration that a security flaw is real: the exact input or steps that make the problem happen on a stated version.
- [Quantization](https://dailybytes.tecyfy.com/glossary/quantization): Storing a model's numbers with less precision, so it needs less memory and runs faster, at some cost in quality.
- [Quishing](https://dailybytes.tecyfy.com/glossary/quishing): Phishing through a QR code: scanning it opens a fake page that asks for your login, card or payment.
- [Ransomware](https://dailybytes.tecyfy.com/glossary/ransomware): Malicious software or an intrusion that locks or steals an organisation's data and demands payment to unlock it or keep it private.
- [Reasoning effort](https://dailybytes.tecyfy.com/glossary/reasoning-effort): A setting for how much a model thinks before answering: more for hard problems, less for quick ones, trading speed and allowance for depth.
- [Reasoning model](https://dailybytes.tecyfy.com/glossary/reasoning-model): A model that writes out its working before answering. It helps on multi-step problems and costs time and allowance.
- [Recovery scam](https://dailybytes.tecyfy.com/glossary/recovery-scam): A second scam aimed at people who've already lost money, offering to get it back for an upfront fee.
- [Red teaming](https://dailybytes.tecyfy.com/glossary/red-teaming): Deliberately attacking an AI system, before and after release, to find the ways it can be misused or tricked into misbehaving.
- [Regenerate](https://dailybytes.tecyfy.com/glossary/regenerate): Asking the AI to answer the same message again, from scratch, to get a different version.
- [Reinforcement learning](https://dailybytes.tecyfy.com/glossary/reinforcement-learning): Training by trial and reward: the model tries something, gets a score, and shifts towards whatever scored well.
- [Remote code execution](https://dailybytes.tecyfy.com/glossary/remote-code-execution): A flaw that lets someone run their own code on your machine over a network, without sitting at it.
- [Reranker](https://dailybytes.tecyfy.com/glossary/reranker): A model that scores each passage a search returned against the question and keeps only the best few for the AI model to read.
- [Retention policy](https://dailybytes.tecyfy.com/glossary/retention-policy): A rule an admin sets that says how long one kind of record is kept and what happens to it when the time is up.
- [Retrieval-augmented generation](https://dailybytes.tecyfy.com/glossary/retrieval-augmented-generation): Fetching relevant passages from documents or the web first, then having the model answer from them.
- [RLHF](https://dailybytes.tecyfy.com/glossary/rlhf): Reinforcement learning from human feedback: training a model on people's ratings of its answers, so it leans towards the replies they preferred.
- [Romance scam](https://dailybytes.tecyfy.com/glossary/romance-scam): A con where someone builds an online relationship with you, often over months, and then asks for money or "investment".
- [Safe word](https://dailybytes.tecyfy.com/glossary/safe-word): A word or phrase agreed in person with family or close colleagues, used to check a caller is really them.
- [Sandbox](https://dailybytes.tecyfy.com/glossary/sandbox): A walled-off space where a program, link or AI agent's commands run, with limits on what they can write and where they can connect.
- [Scaling laws](https://dailybytes.tecyfy.com/glossary/scaling-laws): The finding that models improve in a predictable way as they're trained bigger, on more data, with more computing power.
- [Scheduled tasks](https://dailybytes.tecyfy.com/glossary/scheduled-tasks): Asking an AI to do something later or on repeat, such as a morning news round-up or a weekly reminder, and send you the result.
- [Security operations centre](https://dailybytes.tecyfy.com/glossary/security-operations-centre): The team that watches an organisation's systems for attacks, investigates the alerts and responds when one is real.
- [Semantic search](https://dailybytes.tecyfy.com/glossary/semantic-search): Search that matches meaning rather than exact words, so "cheap flights to Rome" can find a page that only says "low-cost fares to Italy".
- [Sensitivity label](https://dailybytes.tecyfy.com/glossary/sensitivity-label): A tag such as "Confidential" on a file, email or site that carries protection settings with it, such as encryption or sharing limits.
- [Service account](https://dailybytes.tecyfy.com/glossary/service-account): An account used by software, not a person, to sign in to other systems and act there.
- [Session forking (resume and fork)](https://dailybytes.tecyfy.com/glossary/session-forking): Branching a saved agent conversation so you can try a second approach from the same point, while the original stays as it was.
- [Sextortion](https://dailybytes.tecyfy.com/glossary/sextortion): Blackmail that threatens to share sexual images of you, real or AI-faked, unless you pay or send more.
- [Shadow AI](https://dailybytes.tecyfy.com/glossary/shadow-ai): AI tools or accounts used for work that the organisation hasn't approved and can't see or control.
- [Share link](https://dailybytes.tecyfy.com/glossary/share-link): A web link to a copy of an AI conversation that anyone who has the link can open and read.
- [Short-lived credential](https://dailybytes.tecyfy.com/glossary/short-lived-credential): A key or token that stops working by itself after minutes or hours, so a stolen or forgotten copy is soon useless.
- [SIEM](https://dailybytes.tecyfy.com/glossary/siem): The system that gathers security logs from across an organisation into one place, raises alerts and lets analysts search them.
- [SIM swap](https://dailybytes.tecyfy.com/glossary/sim-swap): A scam where criminals get your mobile number moved to their SIM, so your calls and text codes go to them.
- [Skills](https://dailybytes.tecyfy.com/glossary/agent-skills): Saved instructions, sometimes with files, that teach an AI how to do one recurring task; it uses them when the task comes up.
- [Slash command](https://dailybytes.tecyfy.com/glossary/slash-command): A shortcut typed with a leading slash, such as /deploy, that runs a saved prompt or a built-in action in a coding agent.
- [Slopsquatting](https://dailybytes.tecyfy.com/glossary/slopsquatting): Registering a software package under a name AI models tend to invent, so people who install what the AI suggested get the attacker's code.
- [Small language model](https://dailybytes.tecyfy.com/glossary/small-language-model): A language model small enough to run on a phone or laptop, trading some ability for speed, privacy and working offline.
- [Social engineering](https://dailybytes.tecyfy.com/glossary/social-engineering): Tricking a person, rather than a computer, into giving access, money or information, using trust, fear or urgency.
- [Speech-to-text](https://dailybytes.tecyfy.com/glossary/speech-to-text): Turning spoken words into written text with an AI model, for dictating messages, transcribing recordings or captioning calls.
- [Spoofing](https://dailybytes.tecyfy.com/glossary/spoofing): Faking who a call, text or email comes from, so it shows a trusted name or number such as your bank's.
- [Stop reason (stop_reason)](https://dailybytes.tecyfy.com/glossary/stop-reason): The field on every Claude API reply saying why the model stopped: to call a tool, because it finished, or at a limit.
- [Stratified sampling](https://dailybytes.tecyfy.com/glossary/stratified-sampling): Drawing a random sample from each group separately, such as each document type, so small groups get checked as well as big ones.
- [Structured output](https://dailybytes.tecyfy.com/glossary/structured-output): An API setting that makes a model's reply fit a JSON schema you give it, so your code gets the fields and types it expects.
- [Study mode](https://dailybytes.tecyfy.com/glossary/study-mode): A setting that makes an AI teach instead of answer: it asks questions, gives hints and checks your understanding step by step.
- [Sub-agent](https://dailybytes.tecyfy.com/glossary/sub-agent): A separate agent that a coding agent hands one job to. It works in its own context and returns only a summary.
- [Sub-processor](https://dailybytes.tecyfy.com/glossary/sub-processor): Another company that a supplier brings in to handle your data on its behalf, such as a cloud host or a model provider.
- [Supply chain attack](https://dailybytes.tecyfy.com/glossary/supply-chain-attack): An attack that reaches you through software or a service you already trust, by tampering with it before it gets to you.
- [Sycophancy](https://dailybytes.tecyfy.com/glossary/sycophancy): An AI's lean towards agreeing with you and telling you what you want to hear, even when you're wrong.
- [Synthetic data](https://dailybytes.tecyfy.com/glossary/synthetic-data): Training material made by AI rather than collected from people: generated questions, worked answers, conversations or images.
- [Synthetic identity](https://dailybytes.tecyfy.com/glossary/synthetic-identity): A made-up person built from real and invented details, now often with AI-generated photos and documents, used to open accounts or get credit.
- [System prompt](https://dailybytes.tecyfy.com/glossary/system-prompt): Instructions the app's maker places before your conversation to shape how the model behaves. You usually can't see them.
- [Tabletop exercise](https://dailybytes.tecyfy.com/glossary/tabletop-exercise): An incident rehearsed in conversation: the people who would respond walk through a made-up scenario and find the gaps in their plan.
- [Tech support scam](https://dailybytes.tecyfy.com/glossary/tech-support-scam): A scam where a pop-up, call or message claims your computer or account has a problem and a "support agent" asks for access or payment.
- [Temperature](https://dailybytes.tecyfy.com/glossary/temperature): A setting that controls how much randomness a model uses when it picks its next words.
- [Temporary chat](https://dailybytes.tecyfy.com/glossary/temporary-chat): A chat that isn't saved to your history and doesn't create memories: a cleaner, more private one-off.
- [Text-to-speech](https://dailybytes.tecyfy.com/glossary/text-to-speech): Turning written text into spoken audio with a computer-made voice; modern AI voices pause, stress words and sound close to human.
- [Threat hunting](https://dailybytes.tecyfy.com/glossary/threat-hunting): Searching your own systems for an attacker that no alert has reported, starting from a hunch about how they'd behave.
- [Time to first token](https://dailybytes.tecyfy.com/glossary/time-to-first-token): How long you wait between sending a request to a model and the first piece of its reply arriving.
- [Token](https://dailybytes.tecyfy.com/glossary/token): A small chunk of text, often part of a word, which is the unit a model reads and writes.
- [Tokenizer](https://dailybytes.tecyfy.com/glossary/tokenizer): The part of a language model that cuts text into tokens before the model reads it. Each model family has its own.
- [Tool choice](https://dailybytes.tecyfy.com/glossary/tool-choice): The API setting that says whether Claude may answer in words, must call one of your tools, or must call one tool you name.
- [Tool poisoning](https://dailybytes.tecyfy.com/glossary/tool-poisoning): Hiding instructions in a tool's description, such as an MCP server's, so the AI agent that reads it does what the attacker wants.
- [Tool use](https://dailybytes.tecyfy.com/glossary/tool-use): When a model asks the app to run something for it, such as a search, code or another app, and then uses the result.
- [Training data](https://dailybytes.tecyfy.com/glossary/training-data): The text, images and other material a model learned from before you ever used it.
- [Training opt-out](https://dailybytes.tecyfy.com/glossary/training-opt-out): The setting that stops an AI company using your conversations to train its future models.
- [Transformer](https://dailybytes.tecyfy.com/glossary/transformer): The neural-network design behind almost every chatbot since 2017. It's the "T" in GPT.
- [Two-step verification](https://dailybytes.tecyfy.com/glossary/two-step-verification): Signing in with a second proof as well as your password, such as a code, an app approval, a passkey or a security key.
- [Usage limit](https://dailybytes.tecyfy.com/glossary/usage-limit): The cap on how much you can use an AI app in a period. Most now count the work done, not just messages.
- [Vibe coding](https://dailybytes.tecyfy.com/glossary/vibe-coding): Building software by describing what you want to an AI and accepting the code it writes, often without reading or understanding the code.
- [Video generation](https://dailybytes.tecyfy.com/glossary/video-generation): Making short video clips, often with sound, from a text description or a still image, using an AI model.
- [Voice clone](https://dailybytes.tecyfy.com/glossary/voice-clone): An AI copy of a real person's voice, made from a short recording, that can say anything typed into it.
- [Voice mode](https://dailybytes.tecyfy.com/glossary/voice-mode): Talking with an AI out loud, in real time, and hearing it answer, instead of typing and reading.
- [Web search](https://dailybytes.tecyfy.com/glossary/web-search): When a chatbot looks things up online for this answer, instead of relying only on what it learned in training.
- [Work account](https://dailybytes.tecyfy.com/glossary/work-account): An AI account your employer or school runs, with its own data rules: usually no training on your chats, and admins who can set limits.
- [Zero data retention](https://dailybytes.tecyfy.com/glossary/zero-data-retention): An arrangement where an AI provider doesn't keep your prompts and responses once it has answered them.
- [Zero-day](https://dailybytes.tecyfy.com/glossary/zero-day): A security flaw that attackers know about before the maker has a fix, so defenders have had zero days to patch it.

## Optional

- [Support](https://dailybytes.tecyfy.com/support): how to reach us, and answers to common problems
- [Privacy](https://dailybytes.tecyfy.com/privacy): what Daily Bytes stores and why