跳到主要内容

节选或转载版权声明

以下提供两个可直接套用的 Markdown 示例：一个是未授权的节选/摘要（不整篇转载），一个是已获授权的整篇转载。

示例 A：未授权（仅节选/摘要）​

# [Excerpt] The Ethics of Generative AI

**Author:** John Smith

**Source:** TechReview

**Original link:** https://www.techreview.example.com/the-ethics-of-generative-ai

**Note:** All rights belong to the original author. This post includes brief excerpts for commentary and review. Please read the full article at the source.

**Published** on 2025-10-13

## Commentary

In this post I highlight three ethical dilemmas the author raises and add some notes on mitigation.

## Excerpt

> “Generative AI systems inherit biases from their training data...”
>
> “Transparent model documentation is crucial for accountability...”

## My notes

- The author’s second point on dataset audits aligns with recent OECD guidance.
- Practical mitigation includes continuous bias testing and public model cards.

（SEO：节选/摘要不是重复全文，一般 canonical 指向你自己的页面）

<head>
<link rel="canonical" href="https://your-site.example.com/posts/ethics-of-genai-excerpt" />
</head>

示例 B：已获授权的整篇转载​

# [Reprinted] The Ethics of Generative AI

**Author:** John Smith
**Originally published at:** TechReview (https://www.techreview.example.com/the-ethics-of-generative-ai)
**Original date:** 2024-08-10
**Note:** Republished with permission.
**License:** All rights reserved by TechReview.

> Originally published at TechReview on 2024-08-10. Republished here with permission from the publisher and the author. Minor formatting edits only.

## Full text

（以下为原文全文，按授权约定保留署名与图片版权信息）

- Paragraph 1...
- Paragraph 2...
- ...

（SEO：整篇授权转载时，将 canonical 指向原文以避免重复内容冲突）

<head>
<link rel="canonical" href="https://www.techreview.example.com/the-ethics-of-generative-ai" />
</head>