The AI-social media dilemma: Can TikTok, Meta, and diVine curb low-quality content without hurting creators?

Social platforms are grappling with a difficult balancing act. Generative AI can deliver a revolutionary engagement opportunity by turning even more consumers into creators. However, low-quality AI content also poses a threat to the user experience of social platforms. TikTok, Meta, and diVine have all introduced new tools to protect audiences from the worst of AI while ensuring creators can benefit from the best it has to offer. The question is whether social platforms can clamp down on low-quality AI content without catching good-quality AI content in the crosshairs.

What has caught some parts of the industry off guard is how quickly AI content is proliferating. Too much ‘AI slop’ will cause audiences to disengage. MIDiA has championed AI’s creator economy potential in its research (The State of the Video Creator Economy and The future of creator apps reports) while also warning of its threats in Generative AI | The double-edged sword for social video platforms and content creators.

So far, the interventions appear to lack nuance. TikTok's new tool will let users dial down how much generative AI content appears in their feed, or in some cases, increase it if they wish (per Business Insider). TikTok can take such steps because it requires creators to label content produced with AI tools.

Similarly, Meta is using a new ‘content protection tool’ to help creators weed out AI content attempting to duplicate and monetise their ideas, enabling them to track and remove content based on performance.

Meanwhile, Twitter founder Jack Dorsey is introducing a new short-form app – diVine – promising a “certified organic” alternative to the rise of generative AI content with a ‘human first’ focus.

While each approach has its merits, they also risk restricting good-quality AI content or struggling to have long-term impact.

Here are three reasons why:

 1. Will high-quality AI creators be punished?

TikTok’s filter system echoes MIDiA’s call for separate content lanes for generative AI. However, such filters provide a catch-all solution that will reduce both low- and high-quality AI content. The industries remain in the foothills of AI’s potential. Failing to enable good AI content to reach audiences will disincentivise creator engagement, harming tool companies, social platforms, and the wider creator economy. Only by encouraging AI tool adoption and enabling reach will consumer creators grow at pace.

2. Another management time sink

Meta’s tool will help creators protect their IP but risks creating another workflow that eats into time spent producing. Such a tool is helpful for teams but may overwhelm one-man-band creators if it becomes a daily game of whack-a-mole. To succeed, the process needs to be streamlined and automated.

3. Drawing the line on human versus AI content

diVine’s approach may become hard to maintain. Does a ban on AI-created content also extend to AI-assisted content? Using AI to generate a video differs from using AI auto-focus on a camera. Both use AI, but the latter is not autopiloting creative production. If AI-assisted tools are excluded, must a video editor manually cut long-form videos into shorts or risk losing “creative purity”?

Social platforms should reduce low-quality AI content flooding feeds, but they must recognise the challenge ahead. Soon AI features will be so ingrained in workflows that creators may use AI without realising it. The answer may lie in educating audiences on the nuances of AI creation. Equally, creators pushing the frontiers of generative AI should have the platform to do so without fear of demonetisation or downgrading. For the creator economy to realise its growth potential, it must find a way to accommodate both human and AI creation – and everything in between. Blunt instruments will only deliver blunt results.