【OpenAI发布专用AI检测工具，识别DALL-E生成图像准确率高达98%】
美国科技巨头OpenAI于本周二宣布，为应对日益增长的AI生成图像应用，他们已开发出一款专门的图像检测分类器，旨在准确鉴别图像是否由其先进的“文生图”工具DALL·E 3创造。这一工具的发布，旨在帮助研究人员和公众区分真实照片与由人工智能技术合成的图像。
据OpenAI透露，该图像检测分类器在识别DALL-E生成的图像时，准确率达到了惊人的98%。这意味着在大多数情况下，该工具能有效地区分出由DALL·E 3创作的艺术作品和实际拍摄的照片，从而为学术研究、内容审核和版权保护等领域提供强有力的支持。
自周二起，OpenAI已开始接受来自全球研究人员的申请，以获取访问这款检测工具的权限。这一举措表明，OpenAI正积极采取行动，以负责任的态度推动AI技术的发展，并确保其创新成果在透明和可验证的环境中被使用。
随着人工智能在图像生成领域的进步，真实与虚拟之间的界限日益模糊，OpenAI的这一新工具无疑为辨别和理解这些变化提供了一种重要的工具。这不仅有助于保护用户免受误导，也将促进AI技术在媒体、艺术和科学研究等领域的负责任应用。
英语如下：
**News Title:** “OpenAI Launches AI Detection Tool with 98% Accuracy in Identifying DALL-E Created Images”
**Keywords:** OpenAI, DALL-E Detection, Image Recognition
**News Content:**
**OpenAI Unveils Dedicated AI Detection Tool with 98% Accuracy in Identifying Images from DALL·E 3**
American tech giant OpenAI announced on Tuesday that it has developed a specialized image classifier to accurately discern whether images are generated by its advanced text-to-image tool, DALL·E 3. The tool’s release aims to assist researchers and the public in differentiating real photographs from those synthesized by artificial intelligence.
According to OpenAI, the image detection classifier boasts an impressive 98% accuracy rate in identifying images produced by DALL-E. This high level of precision enables the tool to effectively distinguish artwork created by DALL·E 3 from actual photographs, providing robust support for areas such as academic research, content moderation, and copyright protection.
Since Tuesday, OpenAI has started accepting applications from researchers worldwide for access to the detection tool. This move demonstrates the company’s proactive approach to advancing AI technology responsibly, ensuring its innovations are used in a transparent and verifiable context.
As artificial intelligence progresses in image generation, the line between reality and virtuality is becoming increasingly blurred. OpenAI’s new tool thus provides a crucial instrument for discerning and understanding these changes. Not only does it help protect users from misinformation, but it also fosters responsible usage of AI technology in sectors like media, art, and scientific research.
【来源】https://www.cls.cn/detail/1669768