On Tuesday, OpenAI announced a dual approach to combat the growing difficulty in distinguishing authentic images from those generated by artificial intelligence. As AI image generation tools become more sophisticated and widespread, verifying the origin of digital content has become increasingly critical. To address this challenge, the company is implementing two complementary standards designed to provide clear signals of AI involvement while making it difficult for bad actors to erase these markers. First, OpenAI has committed to adopting the C2PA standard, an open protocol developed by the Coalition for Content Provenance and Authenticity. Founded in 2021, this non-profit organization aims to mitigate the harmful effects of AI imagery on public discourse. The C2PA standard embeds visible metadata into images, explicitly stating when they have been created by AI. While this signal is clear and useful for trusted users, it can potentially be manipulated or removed. To provide a more robust layer of security, OpenAI has partnered with Google to integrate SynthID. Developed by Google, this technology places an invisible watermark directly into the image pixels. Unlike metadata, SynthID is designed to persist even when images undergo significant transformations, such as screenshots, resizing, or digital manipulation. OpenAI and Google argue that these two systems address each other's weaknesses; while metadata provides rich provenance information, the invisible watermark offers greater durability against removal attempts. In addition to these technical measures, OpenAI is previewing a public verification tool. Initially, this tool will only check for signals in images generated by OpenAI products, allowing users to easily test whether a file is AI-created. The company expressed hope to expand the tool to cover images from other AI platforms in the future. It is important to note that these protections currently apply exclusively to content generated by OpenAI products. They will not affect the vast volume of imagery produced by other, potentially less reputable AI tools. However, OpenAI states that these measures ensure it does not contribute to the problem of misleading AI-generated content. By adopting C2PA and SynthID, the company aims to make provenance more resilient and foster greater trust in digital media.