OpenAI is enhancing its ability to identify if an image is real or AI-generated by adding new tools to detect images created with its DALL-E generator. The company is also introducing tamper-resistant watermarking tools to assess the likelihood that content originated from generative models. OpenAI has joined the C2PA Steering Committee, a standard for digital content certification, to prove the source of content. It is adding C2PA metadata to images created by DALL-E 3, ChatGPT, and the OpenAI API, as well as integrating it into Sora, their video generation model. OpenAI, in partnership with Microsoft, has launched a $2 million societal resilience fund to support AI education through organizations like Older Adults Technology Services and Partnership on AI. Additionally, OpenAI is developing new provenance methods like watermarking digital content with invisible signals and detection classifiers to resist attempts at removing signals about the content’s origin. The company has opened applications for access to its image detection classifier to the first group of testers.
Source link
Source link: https://timesofindia.indiatimes.com/technology/tech-tips/openais-new-tool-to-help-detect-ai-generated-images-all-details/amp_articleshow/109938002.cms
GIPHY App Key not set. Please check settings