Social media platforms that once promised connection and community are rapidly becoming distorted, impersonal and exhausting — and a growing wave of AI-generated “slop” and deepfakes is accelerating that decline.

For years, sites like Facebook, Instagram and TikTok have been criticized for distorting reality with overly polished lifestyles and unattainable beauty standards. But with today’s generative AI tools — OpenAI’s Sora, Google’s Veo and Midjourney among them — the problem has escalated dramatically. With a few words, anyone can generate hyper-realistic videos, fake people, impossible scenarios and entirely artificial lives.

The result, experts say, is a flood of deceptive content that is eroding trust and connection online.

The term AI slop describes the overwhelming volume of low-quality, AI-generated videos and images now dominating social feeds — from strangely humanlike animals to surreal pranks and chaotic stunts that defy physics. This material is designed to grab attention but leaves users confused, numb or detached.

At the same time, high-quality deepfakes of public figures — including fabricated speeches and staged events — have become disturbingly common.

Filtering through this artificial noise requires more vigilance than ever, and users are increasingly fatigued.

Experts say this isn’t an accident.
“The cynical answer is that social media is now aimed at keeping you connected to the tool, rather than to each other,” said Alexios Mantzarlis, director of Cornell Tech’s Security, Trust and Safety Initiative.

Platforms are eager to highlight AI capabilities for investors and advertisers — often at the cost of user experience. Even before the AI surge, feeds on Instagram and TikTok shifted away from friends and family and toward algorithmically boosted creators. Now, AI-generated influencers, AI vacations and AI “perfect lives” are crowding out human posts even more.

This creates rising insecurity, confusion and a sense of isolation for users seeking authenticity.

“Before, we had unrealistic body expectations,” Mantzarlis said. “Now we’re facing unreal body expectations.”

Platforms including Meta, TikTok and Pinterest have introduced AI labels and promised to restrict harmful deepfakes. TikTok recently began testing a control that lets users limit the amount of AI-generated content in their feeds.

But without strong government regulation — hampered by political disagreements, lobbying and rapid technological change — companies largely police themselves. And many of their efforts so far have been inconsistent or ineffective.

Research shows that simply suspecting content is AI-generated affects trust. A Raptive study found:

- 48% of people find content less trustworthy if they think it might be AI

- 60% feel a weaker emotional connection to AI-generated posts

That distrust is magnified in an era dominated by influencers and algorithmic feeds.

Some experts argue that AI can expand creativity and make content creation more accessible. Tools that simplify production can help new creators thrive.

But they also warn that generative AI can amplify polarization, distort reality and accelerate misinformation — especially in echo-chamber environments.

Social platforms already reinforce biases and spread falsehoods at high speed. AI’s ability to create convincing fake realities could make these problems far worse.

As AI-generated posts blend seamlessly with real ones, many users feel they’re losing control over what they see. A growing number want the ability to filter or eliminate AI-generated content entirely.

For now, that remains limited.

As platforms push AI-driven features and investors demand growth, the human purpose of social media — to connect people to one another — feels increasingly overshadowed.

And unless stronger regulation and better platform controls emerge soon, the artificially crafted reality overtaking our feeds may only continue to grow.