Platforms Expand Controls for AI-Generated Content

Black Enterprise reports that YouTube announced on May 27 that labels for photorealistic or materially AI-generated or altered content would become more prominent. The labels appear beneath long-form videos and as overlays on Shorts.

The outlet also reports that YouTube began rolling out internal detection systems in May that can automatically add labels where it detects significant photorealistic AI use that a creator did not disclose. Separately, Business Insider reported that YouTube's July 20 trust and safety update emphasized rewarding AI use that enhances original storytelling rather than repetitive or manipulative content intended to game the platform.

A YouTube spokesperson told Business Insider that this was a clarification of the platform's existing "inauthentic content" policy, not a new policy. Business Insider also reported that YouTube removed more than a dozen channels in January that had accumulated millions of views with AI-generated videos.

According to Black Enterprise, LinkedIn introduced an option on July 30 for users to flag posts that "seem like AI slop," the platform's term for low-effort AI-generated content without original perspective. LinkedIn Chief Product Officer Hari Srinivasan said more than one million people had used the feature and that members were seeing 40% fewer views of content LinkedIn classified as AI slop than several weeks earlier, Black Enterprise reported.

Substack added a Pangram-powered tool that estimates whether eligible posts, notes, comments, and replies were written by a human or with AI assistance, Black Enterprise reported. The tool applies to content published on or after July 21, and publishers can disable detection on individual posts.

The reported approaches illustrate a broader platform-governance pattern: disclosure tools, automated classifiers, and user feedback each address different parts of the generative-content problem. For practitioners building publishing workflows, detection-based labels should be treated as platform policy controls rather than definitive proof of authorship, particularly where systems are designed to estimate AI assistance rather than establish it conclusively.

Black Enterprise cited brand and AI strategist Amy Zwagerman, who argued that organizations should publish AI-use policies covering human review, confidential information, and customer disclosure. That guidance is an opinion from Zwagerman, not a stated requirement from the platforms.