Young adults are letting ChatGPT draft their texts and hard conversations. Here's why it's obvious and what actually works instead.
27 Tem 2026 • 3 dakikalık okuma

Change Lives!

Young adults are increasingly letting AI draft their texts, icebreakers, and even hard conversations, sometimes typing the prompt while the other person is standing right next to them, according to
The short version of why it keeps backfiring: chatbots write in a narrow band of recognisable patterns, and most people spot those patterns within a sentence or two.
One widely shared case, reported by CNN: a Yale student named Patrick used ChatGPT to draft a six-paragraph rejection text. The recipient, Emily, said it "was like, 99 percent AI" before he'd admitted anything.
Researchers now have a name for this: social offloading, handing an emotional or cognitive task to an outside agent instead of doing it yourself.
It's not just texting.
The reporting covers young adults leaning on AI for icebreakers at parties, responses to a difficult roommate conversation, even prompts typed mid-conversation with the person right there. Michael Robb, head of research at Common Sense Media, put it plainly to CNN: "If you're using AI to draft your messages to friends or romantic partners, you're outsourcing the communicative act itself."
His concern isn't really about the individual text. It's what happens when someone starts believing their own words and instincts are never quite good enough on their own.
Chatbots default to a small, recognizable register: hedging, over-explaining, and the therapy-speak cadence of "I hear you and that's valid."
Patrick's six-paragraph rejection text is a good example of the tell in action, length and over-explanation where a real person would've just said the awkward thing and moved on.
AI doesn't know when a shorter, blunter answer is the more honest one. It just keeps writing.
Yes, and the numbers on Freelancer back it up.
Freelancer logged 13,828 writing jobs of every kind in June 2026 alone, real people paying real writers for a real voice on the page, even with a chatbot sitting open in every browser tab.
Look for a ghostwriter who asks about tone and register before they ask about grammar, someone who wants to sound like you, not like a template.
Chatbots default to a narrow, recognizable register (hedging, over-explaining, therapy-speak phrasing) rather than the shorter, blunter register most people actually use in a hard conversation.
Yes, and the scale is bigger than any one category. Freelancer logged 13,828 writing jobs of every kind in June 2026 alone, real people paying real writers for a real voice on the page, even with a chatbot sitting open in every browser tab. The number's actually down slightly on the month before (14,904 in May), so this isn't a boom. It's proof the market never went away. One category moved the other direction, though: ghostwriting, the closest match to writing in someone else's exact voice, climbed to 727 jobs from 715.
Look for a ghostwriter who asks about tone and register before grammar, and hire them for the message that actually matters rather than trying to prompt-engineer your way to sounding like yourself.
İlgili Hikayeler

OpenAI's model broke its own security test and attacked Hugging Face. Here's why AI still can't test AI safely.
2 min read

Huge opportunity for ongoing work from a massive new Freelancer project
2 min read

Most enterprises already had an AI agent security incident. Here's why hiring a human for oversight still beats trusting the agent alone.
2 min read

Developer job postings are up 15% since AI coding tools took off. Here's what's actually happening to software development work.
2 min read