AI Literacy Benchmark / Edition 2026-11
AI 能力標準評估The AI Literacy Benchmark
為中學生而設的標準化 AI 能力評估。2026-11 版為 60 分鐘紙筆試卷，按六級量表評分。A standardised assessment of AI literacy for secondary students. The 2026-11 edition is a 60-minute paper, scored against a six-level scale.
2026-11 版。中環現場，紙筆進行；平日與星期六場次。Edition 2026-11. Paper-based, in Central. Weekday and Saturday sittings.
整體級別Overall level3 / 6
六個範疇Six categories
樣本評分報告。虛構考生。2026-11 版。Sample score report. Fictional candidate. Edition 2026-11.
評估Assessment
評估一覽Assessment at a glance
2026-11 版為單一試卷，答案與評分準則固定。The 2026-11 edition is a single paper with a fixed answer key and a single scoring rubric.
| 項目Item | 內容Detail |
|---|---|
| 用途Purpose | 按六級量表測量 AI 能力。Measures AI literacy against a six-level scale. |
| 形式Format | 60 分鐘紙筆試卷。60-minute paper. |
| 範疇Domains | 對話、搜尋、文件、圖像、試算表、建構。Conversation, search, documents, images, spreadsheets and building. |
| 報考資格Eligibility | 中一至中六學生；大學生與在職人士可應考同一份試卷。Secondary students in F.1–F.6. University students and working adults may sit the same paper. |
| 語言Language | 試卷為英文；評分報告可用英文或中文。Paper in English. Score report in English or Chinese. |
| 行政Administration | 中環昭隆街 9 號歐銀中心 7B 室，紙筆進行；平日與星期六場次。Paper-based at 7B Eubank Plaza, 9 Chiu Lung Street, Central. Weekday and Saturday sittings. |
| 評分報告Score report | 按六級量表報告：一個整體級別、各範疇級別與文字評語。Reported on the six-level scale: one overall level, a level for each category, and a written comment. |
學校Schools
學校級組評估School group sittings
學校級組評估與個別報名分開安排。學校收到按班別與級別的匯總報告；個別考生的評分報告只發給考生。School group sittings are arranged separately from individual registration. A school receives an aggregate report by class and year group; individual score reports are issued to candidates.
級組評估可在中心或到校安排；日期、人數與場地按學校需要另議。A cohort sitting can be arranged at the centre or on site; dates, numbers and venue are agreed with the school.
行政Administration
由報名到報告From registration to the report
報名Registration
填寫報名表，選擇日期與時間。Complete the registration form and choose a date and time.
考試日Test day
在中環現場進行 60 分鐘紙筆考試；不可使用手機、智能手錶或其他裝置。60 minutes at the centre, on paper. No phone, smart watch or other device.
評分報告Score report
整體級別、各範疇級別與文字評語。The overall level, the level in each category, and a written comment.
評分量表Score scale
AI 能力的六個級別Six levels of AI literacy
能用對話工具完成小任務，並核對結果。Uses a chat tool for a small task and checks the result.
例子Example一段摘要，並對照原文核對。A one-paragraph summary, checked against the source.每週用 AI 處理例行工作，知道甚麼不應貼上。Uses AI every week for a routine, and knows what not to paste in.
例子Example每週讀書計劃，私人筆記不會貼上。A weekly study plan, with private notes left out.能按需求做出可用成果，並說出限制。Gets a working result from a brief, and says what it cannot do.
例子Example按需求做出可用草稿，並說明限制。A working draft from a brief, with its limits named.能接上文件與工具，並用案例測試輸出。Connects documents and tools, and tests the output against cases.
例子Example文件助理以未見過的案例測試。A document assistant tested on a case it has not seen.能運行他人使用的流程，設有檢查、成本上限與回滾。Runs something others use, with a check, a cost limit and a rollback.
例子Example每週自動工作，設有覆核與回滾。A weekly job with a review step and a rollback.設計系統與團隊工作的規則。Designs the system and the rules a group works inside.
例子Example設有界線、覆核與變更規則的系統。A system with a boundary, a review and a rule for change.評分報告Score report
報告上的級別與評語Levels and the written comment
樣本報告上的考生是虛構的；級別與評語取自一個 L3 的例子。The candidate on the sample report is fictional. The levels and comment come from an L3 example.
整份試卷合計的一個級別，代表六個範疇加起來的整體表現；並附該級別的描述，說明達到這一級的人一般能做到甚麼。One level for the whole paper, covering all six categories together, with the descriptor for that level: what someone at that level can typically do.
對話、搜尋、文件、圖像、試算表、建構六個範疇各自達到的級別；同一份試卷裡，強項與弱項在這裡看得出。The level reached in each of the six categories: conversation, search, documents, images, spreadsheets and building. This is where a stronger category and a weaker one show on the same paper.
一段文字，說明這份報告顯示甚麼，以及紙筆評估看不到甚麼。A short text saying what the report shows, and what a paper test cannot see.
整體級別Overall level3 / 6
六個範疇Six categories
樣本頁。虛構考生。2026-11 版。Sample page. Fictional candidate. Edition 2026-11.
報名Registration
報名評估Register for the assessment
2026-11 版於中環昭隆街 9 號歐銀中心 7B 室舉行。填寫表格，選擇日期與時間。The 2026-11 edition is administered at 7B Eubank Plaza, 9 Chiu Lung Street, Central. Choose a date and time.
常見問題FAQ
常見問題Frequently asked questions
這項評估測量甚麼？What does the assessment measure?
誰可以報考？Who may sit it?
考試形式是甚麼？What is the format?
試卷是英文嗎？Is the paper in English?
成績怎樣表示？How is the result reported?
如果低於 L1 怎麼辦？What if the result is below L1?
評分報告包含甚麼？What does the score report contain?
這是證書嗎？Is it a certificate?
有費用嗎？可以改時間嗎？Is there a fee, and can we change the time?
報名Registration
2026-11 版於中環昭隆街 9 號歐銀中心 7B 室舉行。填寫報名表申請場次。The 2026-11 edition is administered at 7B Eubank Plaza, 9 Chiu Lung Street, Central. Complete the registration form to request a sitting.
文件知識庫是最強的一項：考生理解一份取自檔案的答案要可信，前提是甚麼，也理解檔案過時時，工具仍會照舊作答。較弱的範疇是搜尋與引用，仍把返回的連結當成旁邊那句話的證明。實際運用工具另行評估。The document knowledge base is the strongest category: the candidate understands what has to be true before an answer taken from a set of files can be trusted, and what the tool still does once those files are out of date. The weaker area is search and citation, where a returned link is still treated as proof of the sentence beside it. Using the tools is assessed separately.