OpenAI is adding text watermarking in ChatGPT and Codex

OpenAI 在 ChatGPT 和 Codex 中加入文本水印功能

An invisible, machine-readable watermark in text output is rolling out to ChatGPT and Codex, but only for users in the European Union at first. 一种不可见、可由机器读取的文本输出水印正在向 ChatGPT 和 Codex 推送，但初期仅限于欧盟地区的用户。

OpenAI says its textGrain watermarking “matched or exceeded” other approaches like Google DeepMind’s SynthID for text, which is also the basis for the watermarking Anthropic announced in August. Like OpenAI, Anthropic made the move to meet the requirements of the EU’s AI Act; however, not everyone was happy to learn about the addition. OpenAI 表示，其 textGrain 水印技术在性能上“达到或超过了”其他方案，例如 Google DeepMind 的 SynthID（这也是 Anthropic 在八月份宣布的水印技术基础）。与 OpenAI 一样，Anthropic 此举是为了满足欧盟《人工智能法案》（EU AI Act）的要求；然而，并非所有人都对这一新增功能感到满意。

OpenAI also included scores from AI benchmarks showing similar performance from watermarked and unwatermarked text. But it notes that textGrain “does not guarantee reliable detection,” and says text watermarks don’t verify accuracy, determine who owns the text, measure how much a human contributed, or prove human authorship. OpenAI 还列出了 AI 基准测试的评分，显示带有水印和不带水印的文本表现相似。但该公司指出，textGrain “不能保证可靠的检测”，并表示文本水印无法验证准确性、确定文本所有权、衡量人类贡献度，也无法证明文本是由人类创作的。

Rolling out text watermarking in the EU. Over the coming weeks, we will introduce text watermarking to eligible ChatGPT and Codex users across all plans in the EU only. We are not making text watermarking a global default at launch. This regional approach gives us room to learn from real-world use and feedback. 在欧盟推出文本水印。 在接下来的几周内，我们将仅向欧盟地区所有计划中符合条件的 ChatGPT 和 Codex 用户引入文本水印。我们不会在发布时将文本水印设为全球默认选项。这种区域性的方法让我们有空间从实际使用情况和反馈中进行学习。

Enabling opt-in watermarking for API customers. Starting today, API customers around the world will be able to opt in to watermarked text outputs for select models. This lets customers decide how watermarking fits their transparency obligations and the experiences they provide to users. We are also working with cloud partners to make watermarking available for OpenAI model outputs accessed through their services in the coming weeks. 为 API 客户启用可选水印。 从今天开始，全球的 API 客户可以选择为其部分模型开启文本输出水印。这让客户能够自行决定水印如何符合其透明度义务以及他们为用户提供的体验。我们还在与云合作伙伴合作，以便在未来几周内，通过其服务访问的 OpenAI 模型输出也能使用水印功能。

Providing detector access to researchers and expert organizations. Approved researchers and expert organizations can apply starting today. In accordance with the Code of Practice, access will be initially granted on a case-by-case basis to support evaluation and improvement of text provenance. The tool will report whether it detects an OpenAI watermark, without identifying the user or revealing their prompts or conversations. Given the risk of missed watermarks and false positives, we are not making it publicly available at launch. 向研究人员和专家组织提供检测器访问权限。 获批的研究人员和专家组织从今天起可以提出申请。根据《实践准则》，访问权限最初将按个案授予，以支持对文本来源的评估和改进。该工具将报告是否检测到 OpenAI 水印，而不会识别用户或泄露其提示词或对话内容。考虑到水印漏检和误报的风险，我们不会在发布时将其公开。