gen‑ai.news
← Back
Multimodal

SynthID Watermarking Gains OpenAI, Nvidia, and Industry Backing

SynthID, Google DeepMind's system for embedding invisible watermarks in AI-generated content, is moving from a Google-proprietary tool toward something closer to an industry standard. At Google I/O, the company announced that OpenAI, Nvidia, and several other AI providers are adopting SynthID, meaning generated images and videos from those platforms will carry markers that can be verified through Google's detection tools.

OpenAI confirmed separately that output from ChatGPT image generation, Codex, and its API will include both SynthID watermarks and C2PA Content Credentials metadata. C2PA works differently from SynthID - it embeds a manifest in the file's metadata that records provenance information, whereas SynthID hides signal in the pixel or audio data itself in a way designed to survive common edits and re-saves. Using both together is a hedge: C2PA is easier to strip but more human-readable, while SynthID is harder to remove but requires specific detection software.

The practical significance is one of scale. SynthID has been applied to Google-generated content for some time, but its value as a detection tool is limited when most AI-generated content online comes from other providers. Bringing OpenAI and Nvidia output under the same watermarking umbrella substantially increases the share of AI content that could, in principle, be identified as such.

Open questions remain. Watermarks embedded at generation time can be degraded by screenshot, re-encoding, or deliberate adversarial attack. And the detection infrastructure - the tools that read SynthID markers - is not yet uniformly integrated into social platforms or browsers where most people encounter AI content. The expansion is a meaningful step, but the gap between technical capability and consumer-facing transparency is still wide.

Read at Ars Technica →
Share:X

Enjoy this story? Get the next one in your inbox.

Twice a week: the most important stories in generative image and video AI, distilled into a 2-minute read.

Free. Unsubscribe any time. No spam, ever.

Your next read

Multimodal

Industry leaders share new perspectives on generative media for startups

Google for Startups has published a new report examining how early-stage companies are approaching generative media tools and workflows. The findings draw on perspectives from founders and industry figures navigating this space. The report aims to offer practical context for startups integrating AI-generated image and video into their products.

Multimodal

Let us filter AI slop, you cowards

Content labels on AI-generated images and videos have become more common across major platforms, but critics argue that labeling alone is not enough. The Verge makes the case that YouTube, Instagram, TikTok, and others should go a step further and give users the ability to actively filter AI-generated content from their feeds. Without that option, labels function more as a disclosure footnote than a meaningful tool for audience control.

Multimodal

DaVinci Resolve 21 Officially Released With New Photo Editing, AI Tools, and Much More

Blackmagic Design has shipped the final release of DaVinci Resolve 21, marking one of the most substantial updates the software has seen. The version adds a dedicated Photo page for still-image editing alongside a set of AI-powered tools spread across the editing, color, audio, and visual effects areas of the application.