Detecting AI writing on social media and blogs

AI writing is everywhere. Emails, LinkedIn posts, YouTube comments, political speeches, even your group chats—in 2025, it's nearly impossible to avoid. And while AI can be helpful, it can also feel uncanny, hollow, and disingenuous. So how do you know when something was written by a bot, not a human?

In this post, we’ll teach you how to detect AI-generated writing using a set of increasingly obvious red flags. By the end, you won’t need an AI detector. You'll just know.

The Overuse of Em Dashes ( — )

Em dashes are a stylistic punctuation mark used to break up thoughts or add emphasis. While human writers use them sparingly, large language models like ChatGPT love them.

Why it matters: Most people don’t even know how to type an em dash (on Mac: Option + Shift + Minus), so if you see them frequently in casual writing, especially from an unexpected source (e.g., YouTube comments, emails), it’s suspicious.

Important note: The presence of an em dash alone doesn't mean AI wrote it. But it's a strong clue when combined with others.

The "It’s Not Just X, It’s Y" Structure

AI loves rhetorical symmetry. The format:

"It's not just about [X]. It's about [Y]."

is everywhere in AI writing. It feels punchy and balanced, but it’s algorithmically overused.

Examples:

• "It's not just about uploading in 4K. It's about making every shot count."

• "AI isn't just a productivity tool. It brings us closer to our mission."

Once you notice this pattern, you’ll start seeing it everywhere.

Lists of Three (or Five)

AI-generated content loves lists of three: they’re persuasive, easy to remember, and emotionally satisfying.

Examples:

• "It’s about telling a story, conveying emotion, and creating a visually compelling image."

• "Secure borders, cheaper bills, more jobs."

Humans like threes too, but AI relies on them like a crutch. When combined with parallel sentence structures, it’s a tell.

Hollow Language and Vague Praise

AI often praises you or your work with oddly generic, corporate language.

Examples:

• "Your travel vlogs are genuinely captivating."

• "Wonderful question! Truly insightful. You’re a visionary."

It sounds flattering but feels empty—because it is.

Weird, Forced Analogies

AI tries to sound profound by using analogies. But the results often feel off.

Examples:

• "Spotting AI is like being a lighthouse in a fog."

• "It’s like a Band-Aid made of sandpaper."

These are neither helpful nor human. They’re trying too hard to impress.

Over-Explanation and Repetition

AI tends to repeat itself, clarify unnecessarily, or provide obvious context.

Example:

• "The color of the sky is a phenomenon that has fascinated people for centuries."

Instead of getting to the point, AI circles it. And then circles it again. If you feel like you’re reading the same sentence three times, you probably are.

The Uncanny Tone

AI often sounds correct, but feels wrong. There's a subtle lack of authenticity. It's hard to describe, but once you've seen enough of it, you can spot it immediately.

Example:

• "Striking the perfect pose in photography isn't just about looking good—it's about telling a story."

Technically fine. Emotionally dead.

Odd Word Choices & Safe Vocabulary

AI overuses buzzwords and safe phrases that humans typically avoid in casual writing.

Examples:

• Innovative

• Delve

• Elevate

• Practical solutions

They sound polished but generic, like the verbal equivalent of stock photos.

Emoji-Bullet Formatting

AI-generated LinkedIn posts and emails often use emojis to punctuate bullet points. This trend started with humans but has now become an AI hallmark.

Example:

🌟 Telling a story

🌟 Conveying emotion

🌟 Creating visuals

Used alone, it’s not a red flag. Paired with the other signs? Suspicious.

Lacks Personal Detail or Tangents

Real people include themselves in their writing. They tell stories, reference personal experience, and go on tangents.

Example of human writing:

• "As a northerner, I’ve never liked jellied eels. Tried them once, hated it."

AI can mimic tone, but rarely adds real, grounded context like this. Humans also go off-topic in natural, interesting ways. AI tends to stay rigidly focused unless prompted to wander.

The Ultimate Test: Vibe Check

If something feels off, it probably is. That "off" feeling? That’s your gut detecting inauthenticity. AI doesn’t have vibes. You do.

So when a comment or post hits that uncanny sweet spot of being too perfect, too structured, and too generic, trust your instincts.

AI has signature tells you can usually find... if you know what you're looking for.

Can You Hear AI Too?

Spotting AI in writing is one thing. Hearing it in speech—especially in YouTube videos or podcasts—is the next level. Listen for overly smooth phrasing, repeated sentence structures, or voiceover-style tonality with no real texture. Real people stumble. Real people tangent. Real people vibe.