10 Ways to Spot AI-Generated Posts Fueling South Africa’s Xenophobia Crisis

As tensions around immigration and xenophobia continue to flare in South Africa, social media platforms are increasingly being flooded with misleading, manipulated, and AI-generated content designed to provoke outrage, drive engagement, and generate advertising revenue.

Many of these posts appear authentic at first glance. They often feature shocking images, emotional videos, or alarming claims about foreigners, crime, jobs, or public services. But in many cases, the content is fabricated, taken out of context, or generated entirely by artificial intelligence.

The spread of such content can inflame tensions, deepen divisions, and make it harder for people to distinguish fact from fiction.

Here are 10 ways to identify potentially misleading AI-generated posts before sharing them.

1. The Post Triggers an Immediate Emotional Reaction

Disinformation is designed to make people angry, fearful, or outraged. If a post instantly provokes a strong emotional response, take a moment to verify it before reacting or sharing.

Manipulators understand that emotion spreads faster than facts.

2. There Is No Credible Source

Many viral posts make dramatic claims but fail to cite reputable news organisations, official government statements, or verifiable evidence.

If a post references “reports,” “sources,” or “witnesses” without naming them, treat it with caution.

3. Images Look Real but Something Feels Off

AI-generated images have improved dramatically, but they still often contain subtle flaws:

• Distorted hands or fingers
• Unnatural facial expressions
• Inconsistent shadows or lighting
• Strange background details
• Blurred text on signs or clothing

Zooming in often reveals inconsistencies.

4. The Video Lacks Context

A genuine video should answer basic questions:

• Where was it filmed?
• When did it happen?
• Who recorded it?

Many viral videos circulating during social tensions are old clips from unrelated events, sometimes from entirely different countries.

5. The Account Is Built Around Outrage

Check the profile posting the content.

Accounts that repeatedly publish inflammatory content about immigration, crime, race, or politics may be optimised for engagement rather than accuracy. Many exist primarily to generate clicks, followers, or advertising revenue.

6. Reverse Image Search Finds Something Different

A simple reverse image search can reveal whether an image has appeared elsewhere before.

Many supposedly “new” images are actually years old or originate from unrelated incidents in other regions.

7. The Numbers Don’t Add Up

Disinformation often includes statistics without evidence.

Claims such as “millions of jobs stolen” or “thousands arrested in one day” should be verified through official data, reputable media reports, or recognised research institutions.

8. The Language Is Designed to Divide

Posts that portray entire groups as criminals, invaders, or threats often rely on generalisations rather than facts.

Be wary of content that blames complex social and economic problems on a single group of people.

9. The Story Appears Nowhere Else

If a shocking incident is genuine, multiple reputable news outlets will usually report it.

If a claim is spreading widely on social media but cannot be found on established news platforms, it may be false or exaggerated.

10. Fact-Check Before You Share

Before reposting content:

• Search for independent confirmation.
• Check reputable news websites.
• Verify images and videos.
• Look for official statements.
• Review fact-checking organisations.

A few minutes of verification can prevent false information from reaching thousands of people.

The Cost of Sharing Misinformation

Artificial intelligence has made it easier than ever to create convincing fake content. Deepfake videos, AI-generated images, synthetic voices, and automated social media accounts can all be used to manipulate public opinion and exploit existing social tensions.

While technology companies and governments continue to develop countermeasures, individual users remain the first line of defence.

In an age where outrage is profitable and misinformation spreads at unprecedented speed, critical thinking may be one of the most important digital skills South Africans can develop.

Before sharing a post that provokes anger or fear, ask a simple question: Is this verified, or is someone trying to manipulate me?