Online platforms label AI content, but most still lack a filter to hide it

In short: Big social and video platforms are adding labels to AI-made content, but most still do not let users filter that content out.

What's going on

Several major platforms now try to label content made with AI, including images, videos, and music. The Verge reports that YouTube, Instagram, TikTok, and Meta’s apps have increased these labeling efforts over the past year.

The basic idea is simple. If something was created by a computer system, it should carry a notice, like a sticker on a package. Some systems do this by adding hidden information to a file, like an invisible tag on a piece of luggage, but that info can be removed.

The Verge asked Meta, Google, TikTok, and Spotify whether they plan to let people filter out content that has these AI labels. TikTok and Spotify did not respond, Google said it had nothing to share, and Meta did not provide a comment that could be quoted. In practice, that means users still mostly cannot tap a simple “no AI” switch.

A few sites do offer limited controls. The Verge points to DeviantArt and Pinterest, where settings can reduce how much AI content you see. But the options are hard to find, and the writer says they still saw many posts that looked AI-made.

What to watch

If platforms add real filters, it could test whether their AI labels are accurate and consistent. It could also raise new questions about mistakes, since Meta and YouTube have previously labeled some human-made posts as AI-made. For now, the main thing to watch is whether platforms turn labeling into something people can actually use, not just something they can read.

Source: The Verge AI