As AI video generation has continued to improve, it has reached the point of becoming oftentimes indistinguishable from reality. This is alarming to anyone who cares about what is real and what is not. We may find ourselves being entertained, amazed, shocked, manipulated, and misled by content that is completely fake. We become further and further removed from reality as content becomes incentivized to be ever more surprising, shocking, amazing, etc. as it competes with other low-effort content. It is no longer easy to tell when something is not real and avoid it. This creates so much distrust as to discourage us from viewing videos altogether.

Therefore we propose a solution to update TikTok, YouTube, and Reels to enable a way to verify content as authentic:

• If a video is created using the in-app camera, perhaps with editing tools allowed but AI tools disabled, create a "verified" badge for the content.
• Continue to allow camera roll uploads, but these videos shall not have the verified badge, or even rather an "unverified" warning.
• Default settings to FILTER OUT unverified content, requiring users to OPT-IN to AI-generated or other unverified (uploaded) content. This way we retain the freedom to view AI and unverified content, but also the freedom to avoid it, and the right to know the difference.
• For platforms with photo uploads, consider a similar feature. Verify in-app camera usage without AI editing tools.
• To allow cross-posting, platforms with this verification feature could collaborate with each other to honor each other's verification, allowing the export of videos from one to another.

In the future we hope similar strategies may be implemented at the device-level as fake content will inevitably have far-reaching consequences on society and legal systems. For now, these updates should be achievable on social media to bring back trust and choice when it comes to the kinds of content we would like to see.