Ofer Familier, Co-Founder and CEO of Dig.
Instagram announced in April 2026 that it would stop recommending content from accounts that primarily repost others’ work. This policy, which previously applied only to Reels, now extends platform-wide to photos and carousels.
The change is part of a broader industry movement. YouTube tied monetization eligibility directly to authentic, original content in mid-2025, and LinkedIn announced measures targeting recycled thought leadership and generic engagement bait. Even on a legislative level, the EU AI Act’s full transparency requirements for AI-generated content come into force in August 2026, reflecting a concern around AI responsibility online.
These are steps in the right direction. Platforms are acknowledging that feeds have become too saturated with inauthentic content to function as reliable signal layers, and they are now intervening structurally. But the layer they are cleaning up is only the one visible on the surface, including reposts, mass-produced AI content and unlabeled synthetic media. The content built to look real from the start remains untouched, and that is the problem brands have yet to reckon with.
The easy layer is not the dangerous one.
There is no doubt that an avalanche of AI-generated content lives on social media. TikTok revealed that over 1.3 billion videos on its platform were labeled as AI-generated. But labeled AI content is only part of the challenge. The greater risk is synthetic content intentionally designed to appear authentic. It can be tailored to resonate with specific audiences, distributed through credible-looking accounts and seeded into communities where it encourages genuine discussion and engagement.
Because this content appears original rather than manipulated, it often falls outside the scope of platform enforcement. Yet it can be used to shape public perception of brands, governments or other organizations. The engagement it generates is real, even if the narrative behind it is manufactured, and recommendation algorithms amplify those authentic interactions without distinguishing how they originated.
That leaves brands with a growing intelligence problem. Social signals increasingly influence product, marketing and strategic decisions, but there is no reliable way to separate organically emerging sentiment from sentiment that has been deliberately engineered. As a result, organizations risk treating manufactured narratives as genuine customer feedback and making business decisions based on distorted signals.
An engineered feed means engineered intelligence.
The consequences of synthetic content lurking on social media are not hypothetical. When Ubisoft launched Star Wars Outlaws in 2024, the game’s commercial underperformance appeared, initially, to reflect genuine customer backlash. Later analysis of the online response revealed a coordinated trolling campaign characterized by an unusual flood of zero-score user reviews, despite solid professional critic scores. Real player grievances were buried under manufactured noise, and the company made product and strategy decisions on the back of a signal it didn’t verify.
The structural risk not addressed by platform enforcement is that engineered content does not tag brands or use traceable keywords. The signals that reveal it, such as account posting histories, coordinated narrative patterns across creators or the gap between a community’s stated profile and what it’s suddenly producing, are invisible to conventional tools. Most brand listening stacks were built for a text-first internet (keyword tracking, mention monitoring and hashtag scraping) before short-form video became a popular medium through which public perception forms and spreads. That architecture was never designed to read the signals that reveal a coordinated campaign, and the gap between what it detects and what is actually shaping perception is where strategic decisions go wrong.
Gartner predicts enterprise spending on battling misinformation and disinformation will surpass $30 billion by 2028, yet most organizations still don’t know how to get a complete picture of the information being disseminated about them, whether it’s real and what it means for them.
Brands have to close the intelligence gap.
Forty-nine percent of U.S. consumers now believe that generative AI has made content quality worse, and brands will need to find a way to combat this negative sentiment toward the content their consumers are seeing online, and fast. AI now lets anyone craft believable synthetic content that walks and talks like the real thing, but the hard work of telling the two apart falls to the brands themselves.
When brands see narratives developing around them on social media, they need to ask more than what their audiences are saying. They need to ask:
• Who said it first?
• Was the buzz built in a community that actually uses the product, or one manipulated to react?
• Does what we’re seeing online match what sales and support are actually hearing?
Answering those questions means tracing where a signal originates, not just measuring its volume. Skip that, and leaders risk making the wrong calls based on signals that were placed rather than earned. It’s no longer a volume game. It’s a question of whether a narrative forming around your brand is authentic. That is certainly hard to do, but necessary.
Social feeds are getting cleaner. The question for brands is whether their intelligence infrastructure is getting smarter. Platforms can flag a repost, but they cannot tell a brand whether the sentiment shaping its next product decision was earned or placed. Knowing that difference can mean a sound business decision or one that won’t hold up in the boardroom.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
