Platforms restrict AI-generated content to prevent a digital race to the bottom

Snapchat will stop recommending entirely AI-generated videos in its Spotlight feed, joining a wave of major platforms moving to curb the spread of automated, low-quality material. The decision by Snap follows similar actions over the past two weeks by YouTube, LinkedIn and Substack. Together, the measures mark a significant shift in how the tech industry is handling the content produced by its own tools.

For platforms that rely on user engagement to sell advertising or subscriptions, the proliferation of AI-generated text, images and videos poses a direct threat to their core business models. The automated flood risks driving away the human creators and active users that attract advertisers. Research indicates that the more fake content users see, the less they trust anything on a platform, undermining its commercial value.

Substack co-founder and chief executive Chris Best recently cited research showing up to 40% of writing on social media is now fake or AI-generated. "It's getting harder to tell what's real on the internet," Best said last week while unveiling a new tool to help readers detect AI writing. He warned that "platforms that reward fakeness will create a race to the bottom."

YouTube updated its monetisation policies this month to cut off revenue for what it calls "inauthentic content." The Google-owned platform will no longer allow creators to make money from videos that are generic, repetitive or template-based—styles easily mass-produced with generative AI. Last year's research found numerous YouTube channels consisting entirely of AI-generated content, some pulling in millions of dollars in revenue and amassing millions of subscribers.

YouTube trust and safety chief Matt Halprin acknowledged that while AI can be used creatively, it is increasingly exploited for scale. "The same technology really enables great stuff," Halprin said. "But it also enables stuff that's kind of content farming, and that's the stuff that we don't want to have."

Professional network LinkedIn is taking a different approach, focusing on the sheer volume of automated spam. The platform introduced a reporting button this week for users to flag suspected AI-generated posts or comments. LinkedIn chief product officer Hari Srinivasan wrote that "AI slop is a top priority for all of us," noting the platform has blocked billions of such attempts in recent months and catches hundreds of thousands daily.

The platforms are careful to draw a line between wholly generated content and tools used to edit or refine human work. Snap will still recommend AI-enhanced content, and LinkedIn is merely removing an automated prompt that offered to "enhance" posts, reverting to a basic proofreading tool. Nonetheless, the collective crackdown signals that the industry now views unpoliced AI content as an economic liability rather than an asset.