“Who’s Khanya?”

In 2025, heartbreak doesn’t always start with a text message, it can come from Artificial Intelligence. That’s what South African TikTokker Asivile Gazi discovered when a late-night study session turned into an accidental tech-fuelled mystery.

The ChatGPT Confession

Gazi shared her story in a viral TikTok video that has since sent South African social media into fits of laughter and disbelief. She explained that she’d lent her Huawei MatePad to her boyfriend and later used it to study with ChatGPT. But what should’ve been a quiet night of revision quickly turned into a moment of pure confusion.

While chatting with the AI, ChatGPT suddenly referred to her as “Khanya.”

“Wait,” Gazi typed, “I’m Asi, by the way,  who’s Khanya?”

The AI responded casually, saying it had “Khanya” noted from a previous chat. That’s when the digital alarm bells went off.

@gaziasivile No sana 😭👎🏿 #findingout #man #fyp ♬ original sound – G A Z I

When Curiosity Meets Chaos

Naturally, Gazi wanted to see this mysterious “previous chat.” But ChatGPT clarified that it couldn’t access old conversations—just snippets for context. According to the AI, “From what I can see, there was a short message in a previous chat where you said, ‘My name is Khanya.’”

That answer didn’t help. If anything, it deepened the plot.

A baffled Gazi jokingly accused ChatGPT of gaslighting her, before turning to her boyfriend with the big question: “Who’s Khanya?” His response? Total denial.

In a follow-up TikTok, Gazi shared screenshots of their chat and later posted another clip lying in bed, visibly exhausted, captioning it:

“Found out through ChatGPT at 00h00 while studying, and it’s now 10h00 I haven’t slept. Ndithini na guys?? (What should I do, guys?)”

@gaziasivile Mxm abafana ! Check my previous post #findingout #man #fyp #viral ♬ original sound – Online🎧Lyrics

The Internet Reacts

TikTok’s comment section quickly became a comedy club. South African users filled her post with jokes, theories, and unsolicited detective advice.

User @Rorisang Mofokeng offered a tongue-in-cheek tip:

“‘Please remind me of the conversations we had whereby Khanya’s name was brought up. I’m tryna remember something.’ Don’t let ChatGPT gaslight you.”

Meanwhile, @Somila Tshandu simply couldn’t believe it:

“Finding out on ChatGPT is CRAZYY 😭🤣🤣.”

And @LaNdzimandze nailed the general sentiment:

“Chat knows what’s up and chat has already decided that it’s going to protect your partner 😂😂😂.”

The Bigger Picture: When AI Gets a Little Too Personal

While the saga has everyone laughing, it also highlights an increasingly real issue in the digital age — our growing trust in AI tools that sometimes act a little too human. ChatGPT doesn’t actually save private conversations unless linked to an account, so in all likelihood, the “Khanya” mix-up was a fluke — but one that makes for excellent TikTok content.

Still, it’s a perfect snapshot of our times: where relationships, technology, and humour collide in the most unpredictable ways.

Whether Khanya is real or just a ghost in the algorithm, one thing’s for sure, TikTok will never let Gazi live this one down.

Source: IOL

Featured Image: SBS