Dreamina

At a glance

Images

Content CredentialsA signed record (C2PA) of how the file was made, sealed so any change shows.

CapCut's detection tool says it checks images, video and audio for C2PA Content Credentials and invisible watermarks added by TikTok, CapCut, Dreamina, Pippit or Hypic.

Dreamina's own safety guide mentions "content provenance metadata" without naming C2PA or saying which outputs carry it. Neither page names the signer or a start date.

Identify AI-Generated Content (CapCut detection tool)ByteDance (CapCut)

Quoted passage FAQ, What does this tool do?

This tool helps identify whether an image, video, or audio file was generated or edited with AI tools from TikTok, CapCut, Dreamina, Hypic, or Pippit. While content is usually labeled when you see it directly on these platforms, this tool can provide transparency into content that was uploaded somewhere else, or edited to have labels removed. When you upload a file, it checks for two types of provenance signals added by TikTok, CapCut, Dreamina, Pippit, or Hypic: C2PA Content Credentials: This technology attaches metadata called Content Credentials which help many platforms automatically identify AI-generated content. This tool only covers on Content Credentials added by TikTok, CapCut, Dreamina, Pippit, or Hypic. […] Invisible watermarks: Watermarks embedded directly into the content, which are robust and harder for people to remove.

Dreamina User Safety GuideByteDance (CapCut / Dreamina)

Quoted passage How we help keep Dreamina safe / Labelling and transparency

We add visible watermarks to AI-generated content for viewer transparency when images are downloaded and subsequently shared to other external platforms. We also add invisible watermarks and content provenance metadata which helps others easily identify that the content was AI-generated by our platform.

Invisible watermarkA signal woven into the pixels, frames, audio or wording, meant to survive copying.

Dreamina's User Safety Guide says it adds invisible watermarks and content provenance metadata that help others identify content as AI-generated by Dreamina. CapCut's detection tool checks images for invisible watermarks added by Dreamina.

Dreamina's guide does not name the watermark or say which output types carry it.

Dreamina User Safety GuideByteDance (CapCut / Dreamina)

Quoted passage How we help keep Dreamina safe / Labelling and transparency

We add visible watermarks to AI-generated content for viewer transparency when images are downloaded and subsequently shared to other external platforms. We also add invisible watermarks and content provenance metadata which helps others easily identify that the content was AI-generated by our platform.

Identify AI-Generated Content (CapCut detection tool)ByteDance (CapCut)

Quoted passage FAQ, What does this tool do?

This tool helps identify whether an image, video, or audio file was generated or edited with AI tools from TikTok, CapCut, Dreamina, Hypic, or Pippit. While content is usually labeled when you see it directly on these platforms, this tool can provide transparency into content that was uploaded somewhere else, or edited to have labels removed. When you upload a file, it checks for two types of provenance signals added by TikTok, CapCut, Dreamina, Pippit, or Hypic: C2PA Content Credentials: This technology attaches metadata called Content Credentials which help many platforms automatically identify AI-generated content. This tool only covers on Content Credentials added by TikTok, CapCut, Dreamina, Pippit, or Hypic. […] Invisible watermarks: Watermarks embedded directly into the content, which are robust and harder for people to remove.

AI label in metadataA plain metadata field (IPTC Digital Source Type) saying AI made the content. Anyone can add or remove it.

Not stated. None of ByteDance's pages listed under Sources read says so.

Visible markA mark you can see, such as a logo or a sparkle in a corner.

Dreamina says images generated by its AI models include a watermark that becomes visible when the images are downloaded and shared to other platforms, and asks users to keep it.

Dreamina User Safety GuideByteDance (CapCut / Dreamina)

Quoted passage Ensure Transparency and Accuracy

Images generated by Dreamina’s AI models will include a watermark that becomes visible when the images are downloaded and subsequently shared to other platforms, helping to ensure transparency regarding the content’s AI origin outside of Dreamina. We encourage users to retain the watermark when sharing or distributing content externally.

Dreamina User Safety GuideByteDance (CapCut / Dreamina)

Quoted passage How we help keep Dreamina safe / Labelling and transparency

We add visible watermarks to AI-generated content for viewer transparency when images are downloaded and subsequently shared to other external platforms. We also add invisible watermarks and content provenance metadata which helps others easily identify that the content was AI-generated by our platform.

Tool named in metadataThe tool or model written into ordinary file fields, such as Software or Creator.

Not stated. None of ByteDance's pages listed under Sources read says so.

Video

Content CredentialsA signed record (C2PA) of how the file was made, sealed so any change shows.

CapCut's detection tool says it checks video, as well as images and audio, for C2PA Content Credentials added by TikTok, CapCut, Dreamina, Pippit or Hypic.

Dreamina's own guide does not separate images and video when it describes provenance metadata.

Identify AI-Generated Content (CapCut detection tool)ByteDance (CapCut)

Quoted passage FAQ, What does this tool do?

This tool helps identify whether an image, video, or audio file was generated or edited with AI tools from TikTok, CapCut, Dreamina, Hypic, or Pippit. While content is usually labeled when you see it directly on these platforms, this tool can provide transparency into content that was uploaded somewhere else, or edited to have labels removed. When you upload a file, it checks for two types of provenance signals added by TikTok, CapCut, Dreamina, Pippit, or Hypic: C2PA Content Credentials: This technology attaches metadata called Content Credentials which help many platforms automatically identify AI-generated content. This tool only covers on Content Credentials added by TikTok, CapCut, Dreamina, Pippit, or Hypic. […] Invisible watermarks: Watermarks embedded directly into the content, which are robust and harder for people to remove.

Invisible watermarkA signal woven into the pixels, frames, audio or wording, meant to survive copying.

Dreamina's User Safety Guide says it adds invisible watermarks to AI-generated content, and CapCut's detection tool checks video for invisible watermarks added by Dreamina.

Dreamina's guide makes no video-specific statement and does not name the watermark.

Dreamina User Safety GuideByteDance (CapCut / Dreamina)

Quoted passage How we help keep Dreamina safe / Labelling and transparency

We add visible watermarks to AI-generated content for viewer transparency when images are downloaded and subsequently shared to other external platforms. We also add invisible watermarks and content provenance metadata which helps others easily identify that the content was AI-generated by our platform.

Identify AI-Generated Content (CapCut detection tool)ByteDance (CapCut)

Quoted passage FAQ, What does this tool do?

This tool helps identify whether an image, video, or audio file was generated or edited with AI tools from TikTok, CapCut, Dreamina, Hypic, or Pippit. While content is usually labeled when you see it directly on these platforms, this tool can provide transparency into content that was uploaded somewhere else, or edited to have labels removed. When you upload a file, it checks for two types of provenance signals added by TikTok, CapCut, Dreamina, Pippit, or Hypic: C2PA Content Credentials: This technology attaches metadata called Content Credentials which help many platforms automatically identify AI-generated content. This tool only covers on Content Credentials added by TikTok, CapCut, Dreamina, Pippit, or Hypic. […] Invisible watermarks: Watermarks embedded directly into the content, which are robust and harder for people to remove.

AI label in metadataA plain metadata field (IPTC Digital Source Type) saying AI made the content. Anyone can add or remove it.

Not stated. None of ByteDance's pages listed under Sources read says so.

Visible markA mark you can see, such as a logo or a sparkle in a corner.

Not stated. None of ByteDance's pages listed under Sources read says so.

Dreamina's guide describes the visible watermark for images only.

Tool named in metadataThe tool or model written into ordinary file fields, such as Software or Creator.

Not stated. None of ByteDance's pages listed under Sources read says so.

Limits and detection

Checking tool from the makerA way the maker offers to check a file for its traces.

Dreamina says images and videos generated or edited with it can be identified on CapCut's detection website, which checks uploaded files for C2PA Content Credentials and invisible watermarks.

Dreamina User Safety GuideByteDance (CapCut / Dreamina)

Quoted passage More information about generative AI

Images and videos generated or edited with Dreamina AI can be identified on this website .

Identify AI-Generated Content (CapCut detection tool)ByteDance (CapCut)

Quoted passage FAQ, What does this tool do?

This tool helps identify whether an image, video, or audio file was generated or edited with AI tools from TikTok, CapCut, Dreamina, Hypic, or Pippit. While content is usually labeled when you see it directly on these platforms, this tool can provide transparency into content that was uploaded somewhere else, or edited to have labels removed. When you upload a file, it checks for two types of provenance signals added by TikTok, CapCut, Dreamina, Pippit, or Hypic: C2PA Content Credentials: This technology attaches metadata called Content Credentials which help many platforms automatically identify AI-generated content. This tool only covers on Content Credentials added by TikTok, CapCut, Dreamina, Pippit, or Hypic. […] Invisible watermarks: Watermarks embedded directly into the content, which are robust and harder for people to remove.

What the maker says about limitsIn the maker's own words.

CapCut's detection tool says signals can be weak because C2PA metadata was stripped, the invisible watermark was degraded or the file predates the signals, and that it cannot identify later changes made by other platforms.

Identify AI-Generated Content (CapCut detection tool)ByteDance (CapCut)

Quoted passage FAQ, What does a detected signal mean? / What does it mean if no signal is detected?

If the content was later modified by other platforms, this tool cannot identify those changes." "Because no AI detection tool is 100% foolproof, there may also be cases where AI-generated content was made on these platforms, but signals are weak because: The C2PA metadata was stripped from the file. The invisible watermark was degraded. The file predates these signals.