이미지: X — 프론티어랩 영상 갈무리
Summary
- On August 20, Anthropic unveiled Claude Academy, a free site teaching Claude products and AI fundamentals
- It is divided into learning paths for five products—Claude.ai, Cowork, Code, Tag, and Platform—and an "AI Fluency" course covering AI concepts
- The flagship course is a 4-hour, 14-lesson class on the 4D framework (Delegation, Description, Discernment, Diligence)
- 공개일
- 2026-08-20 (X, @claudeai)
- 이용 조건
- 무료, 누구나 이용 가능
- 제품 학습 카테고리
- Claude.ai · Claude Cowork · Claude Code · Claude Tag · Claude Platform 5종
- AI Fluency 대표 코스
- Framework & Foundations, 14레슨·퀴즈 1회·총 4시간, 4D 프레임워크(Delegation·Description·Discernment·Diligence)
- 짧은 튜토리얼
- The 4 Properties of AI, 7분
- 능력·한계 코스
- AI Capabilities and Limitations, 13레슨·퀴즈 1회·3.5시간
- 페이지 하단 정책
- 개인정보 처리방침 2026년 7월 8일 시행 명시
Someone who says "I'm not even sure what AI is" and someone who says "I use Claude every day" can now each pick a different course on the same site. On August 20, Anthropic announced via its official X account that Claude Academy had opened. In the post, the company said, "The courses and tutorials are free and available to everyone."
Anthropic is the AI startup that makes Claude. Over the past few months, the company has expanded Claude from a single conversational chatbot into Claude Cowork, which handles entire tasks; Claude Tag, which hands off work by tagging it in team channels; and Claude Platform, which embeds Claude into products via API. In fact, on August 19 and 20, the company continued expanding its developer platform by adding memory, domain control, and a session viewer to Claude Managed Agents. As the product lineup has branched out, guidance on where to start learning has become necessary, and Claude Academy is meant to serve as that entry point.
Five Claude products in one place
The site is broadly split into two tracks. One covers how to use each product. There are five cards: the standard Claude.ai for solving problems and writing documents in a chat window; Claude Cowork, which takes on entire tasks and delivers only the results; Claude Code, for writing code in the terminal, IDE, or browser; Claude Tag, for tagging Claude in team channels to hand off work; and Claude Platform, for embedding Claude into your own products via API, console, and MCP.
The other track is the "AI Fluency" section, focused on understanding AI itself. The flagship course, "AI Fluency: Framework & Foundations," teaches the "4D Framework," made up of Delegation, Description, Discernment, and Diligence. For a quick overview, there's the 7-minute "The 4 Properties of AI," and there's also "AI Capabilities and Limitations," which covers what language models are good and bad at, including next-token prediction, knowledge, working memory, steerability, and context limits.
Course lineup
| Course | Format | Length | Key Content |
|---|---|---|---|
| AI Fluency: Framework & Foundations | Course (14 lessons, 1 quiz) | 4 hours | 4D Framework: Delegation, Description, Discernment, Diligence |
| The 4 Properties of AI | Tutorial | 7 minutes | Summary of AI capabilities and limitations |
| AI Capabilities and Limitations | Course (13 lessons, 1 quiz) | 3.5 hours | Next-token prediction, knowledge, working memory, steerability, context limits |
How to use it
Following the link in the post brings up two buttons on the landing page: "Get started" and "All resources." Here's how to use the site:
- On the landing page, clicking "Get started" shows recommended learning paths, while "All resources" shows the full course list.
- Choose from the top menu among "AI Fluency," "Products," and "Resources."
- Under "Products," click on the product you want to learn among the five cards: Claude.ai, Cowork, Code, Tag, and Platform.
- Under "AI Fluency," selecting a course or tutorial displays the lesson list and estimated time required.
The only stated conditions for use are that it's free and open to everyone; no separate pricing plans or regional restrictions have been announced. As an example of how it might be used, someone new to Claude could start by grasping the concepts through the 7-minute "The 4 Properties of AI." A team evaluating whether to adopt Cowork or Tag could first run the 4-hour AI Fluency course as onboarding material, while a developer building a product via API could check the Claude Platform card for guidance on using MCP and the console.
At the bottom of the page are links to the Privacy Policy, effective July 8, 2026, and the usage policy, allowing users to check how information entered during course use is handled.
Editor's take
The decision to spin off educational content into its own site is itself a signal. Until now, new AI tools were typically introduced with a single release note or blog post, leaving actual usage guidance to be figured out by the community. Rolling out a full 4-hour formal curriculum complete with lessons, quizzes, and estimated completion times is unusual even for this company's product lineup. That suggests Claude has grown from a single chatbot into four distinct branches—Code, Cowork, Tag, and Platform—creating a need for a map to guide users to each one.
For a company of this size, building an in-house training curriculum is ultimately also a way to cut support costs. By setting things up so new users can simply follow a fixed course instead of searching the help center or asking the community, the company can save on the manpower needed to handle inquiries.
If a domestic company is considering adopting Claude, it would be safer to have the relevant staff complete this free course before signing a contract. In particular, for collaborative features like Cowork and Tag that aren't yet familiar to most users, aligning on the concepts before applying them to actual work can reduce trial and error. In the coming weeks, it's likely that the Cowork and Tag courses will be further broken down, or bundled separately as enterprise onboarding packages.
Comments