Meta develops built-in AI Detector to identify synthetic content on its platforms

Meta builds internal AI Detector for synthetic content verification

Meta is developing a built-in AI Detector for its Meta AI chatbot that could allow users to verify whether content was created using artificial intelligence. The feature was first spotted by TestingCatalog, which shared screenshots on X showing a new "AI Detector" menu option appearing directly inside the Meta AI interface 1

1

. The AI detection tool remains inactive on Meta's servers, and clicking the option currently leads to a broken link, indicating the feature is still under development 2

2

.

The discovery was most likely made by enabling feature flags within the app's code, a common method for uncovering unreleased functionality 3

3

. Since the AI Detector cannot be accessed yet, its full capabilities remain unclear. The tool may initially identify AI-generated or enhanced text before expanding to detect images, audio, or video 1

1

. It's also uncertain whether the detector will identify synthetic content from any large language models or only those developed by Meta 3

3

.

Growing pressure for transparency regarding AI-generated material

The feature's development signals growing industry pressure to label AI-generated material as platforms face regulatory demands and user calls for content authenticity 2

2

. Meta has been aggressively pushing generative AI tools across its platforms, including Meta AI features that can quickly generate images and other media. The company is also testing a standalone Vibes app where everything users scroll is AI-generated 1

1

.

These same tools have contributed to a surge of AI slop—low-effort synthetic content spreading across social media feeds. Meta now appears to be building a system that could help identify the very kind of content its technology creates 1

1

. The move mirrors Google, which already offers a similar AI video detection tool in Gemini, suggesting Meta may be following the same direction 1

1

.

Deployment timeline unclear amid separate AI setbacks

Meta has not officially announced the feature, and no release timeline has been disclosed 2

2

. Since the feature was discovered within the code and remains unannounced, it could still change significantly before any public release 1

1

. The development comes as Meta faces separate technical setbacks with its planned large language models, codenamed Avocado, which have experienced delays due to performance issues and gaps with competing models 2

2

3

3

.

If deployed, the internal AI Detector could affect how millions of users verify content authenticity across Meta's services 2

2

. The tool represents a potential shift in how Meta approaches the challenge of distinguishing between human-created and machine-generated material, particularly as regulatory scrutiny intensifies around AI-generated content disclosure. Users should watch for official announcements from Meta regarding the feature's capabilities and whether it will detect AI-generated content from competing platforms or focus solely on Meta's own AI outputs.