Google Gemini更新:允许用户隐藏AI图像水印
用户现在可以生成不带可见水印的图像,但文件元数据中仍保留了不可见的识别标记。

Google为其Gemini平台推出了一项小幅更新,允许用户隐藏AI生成图像上的可见水印。这一变动确保了所生成的内容不再向肉眼显示明显的浮水印。
据原始发布者称,此举旨在为使用该工具创建图像的用户改善视觉体验。此前,此类图像通常会被标注上明显的AI生成属性标识。
尽管移除了视觉标记,但Google明确表示,安全与识别标准仍然保持不变。系统依然会在文件中植入SynthID水印和C2PA元数据。这些标识符被嵌入在图像数据中,但设计上人眼是无法察觉的。
此次更新在为用户提供更简洁、更具可用性的视觉内容,以及保持追踪AI生成媒体的行业标准之间取得了平衡。通过保留技术性标识符,Google旨在确保那些具备读取嵌入式元数据能力的系统仍能验证图像来源。
对于马来西亚的用户和企业而言,这一变化反映了生成式AI工具在向更专业的实用性过渡过程中所发生的持续演变。随着马来西亚各行业日益探索将AI应用于创意和营销工作流程,在保持数字来源可追溯性的同时制作无水印图像,有助于兼顾美学质量与监管透明度。
Source
Originally reported by Cms. Read the original report →
This story was translated from our English report. Read in English →
Join the conversation
We post stories like this all day on Threads. Discuss this story on Threads →
More in AI
World AI Show 2026 Solidifies Malaysia as Regional AI Powerhouse
Kuala Lumpur hosts over 1,500 industry leaders to chart the future of national AI infrastructure and enterprise adoption.

Tupai.ai Pivots From Beta Bot to Peer-Led Math Reform in Schools
With national maths proficiency stalling, this Malaysian startup is scaling a human-centric AI model to bridge the education gap.

Malaysia’s AI Data Centre Expansion Faces Critical Infrastructure Capacity Constraints
Rapid growth in data centre capacity threatens to outpace Malaysia’s power and cooling infrastructure, warns Delta Electronics.

Grafilab Stakes Future on Sovereign AI to Challenge Global Cloud Giants
The local startup aims to leverage data sovereignty and localized compute to secure a foothold in Malaysia’s burgeoning cloud infrastructure market.
