ChatGPT’s new caricature trend has taken over social media, generating exaggerated portraits based on users’ photos and chat history. While many found the results funny, others were left confused or offended by inaccurate depictions. Experts have also flagged privacy concerns, as the trend requires sharing personal data that may be processed by AI systems.

Move over Studio Ghibli, ChatGPT's new caricature trend has exploded across social media. AI-generated portraits that blend personal details into exaggerated, cartoonish depictions are flooding Indian feeds. The simplicity of the prompt - "Create a caricature of me and my job based on everything you know about me" - has helped most users get on board the bandwagon.

Users upload photos and share conversation histories with theis prompt, and ChatGPT spits out images that often hilariously (or horrifyingly) reflect perceived traits, hobbies, or professions. However, not all results are flattering - or accurate. Amid the laughs, privacy concerns have emerged, as the trend requires sharing personal data with AI, potentially exposing sensitive info to OpenAI's systems. H

ere are five instances where the trend veered into disaster territory, leaving users roasted, confused, or outright offended.

1. Crypto trader turned into a food Junkie

Crypto influencer K A L E O (@CryptoKaleo) jumped on the bandwagon, prompting ChatGPT with "make a caricature of me as an American man based on everything you know about me." The result? A rotund figure gleefully devouring burgers, pizza, fries, and shakes amid a fast-food apocalypse. "Am I cooked?" he posted on X.

2. Surprise Makeover

Sports enthusiast Sparkles and JOY (@krissyhaynie) shared her ChatGPT creation with a mix of shock and amusement, "When you ask for a caricature based on what it knows about you 🤣wtf ChatGPT ahahahhaha." The AI delivered a wide-eyed woman with prominent freckles, a bubbly smile, and accessories nodding to her love for teams like the Gamecocks and Hornets. The result felt more like a roast than a tribute, highlighting how AI can amplify traits in unintended, unflattering ways.

3. The educator armed with a chainsaw

Michelle Baker (@MBaker_RunSun), a Director of Differentiation in education, asked for a work-based caricature and got a bizarre first draft - herself wielding a chainsaw amid classroom chaos. "The first attempt was curious! I’ve never used a chain saw so I asked AI what was up with that," she posted on X. This leads to speculation about why AI defaulted to such an odd prop, underscoring the trend's potential for wildly off-base interpretations of professional lives.

4. The fitness guru gender-swapped with a kid

Jeremy Morris (@JMorrisfitness), a reality TV winner and fitness personality, vented frustration after his caricature prompt yielded "a lady with a random kid." "@ChatGPTapp caricatures be like.. gah.. I can’t even explain how bad that is," he tweeted, calling it a 'scam' that only improves with payment. The mismatch - wrong gender, unexplained child- turned what should have been a buff, adventurous depiction into a confusing dud.

5. The crypto researcher as a shady politician

Analytical writer Da Vince (@DaVine95234565) tried the prompt on Grok but shared a ChatGPT-style result that resembled "an unpopular politician" like Nicolas Sarkozy, complete with scheming expression and charts. "Look at what it generated....I weak," he posted, tagging the AI and earning laughs from followers who agreed it looked more like a caricature of a shady figure than a thoughtful researcher.

While the trend's virality shows no signs of slowing, these mishaps serve as a reminder of AI's limitations. Experts warn of privacy risks, as prompts often involve sharing photos or histories that train models further.