Formatting Patterns in AI-Generated Text

Formatting Patterns in AI-Generated Text

Visual and typographical formatting patterns that indicate AI authorship. These patterns often result from AI chatbot default behaviors and prompt engineering instructions.

Title Case in Section Headings

The Pattern

AI strongly tends to capitalize all main words in section headings.

AI tendency:

Global Context: Critical Mineral Demand
Strategic Negotiations and Global Partnerships
High-Stakes Deals: Glencore, China, and Russia

Wikipedia/human convention:

Global context: critical mineral demand
Strategic negotiations and global partnerships
High-stakes deals: Glencore, China, and Russia

Why This Happens

Most AI chatbots are trained on data where headlines and section titles use title case. Wikipedia style prefers sentence case.

Detection

Check if ALL section headings in a document use title case. A mix of styles is more human; consistent title case throughout is an AI signal.

Excessive Boldface

The Pattern

AI may bold various phrases for emphasis in an excessive, mechanical manner, especially:

• Every instance of a chosen word or phrase
• "Key takeaways" style emphasis
• Important terms in every paragraph

Examples

"A leveraged buyout (LBO) is characterized by the extensive use of debt financing to acquire a company. This financing structure enables private equity firms and financial sponsors to control businesses..."

"OKRs (Objectives and Key Results), KPIs (Key Performance Indicators), and visual strategy tools such as the Business Model Canvas (BMC) and Balanced Scorecard (BSC)."

Why This Happens

AI inherits boldface patterns from:

• README files
• Fan wikis and how-tos
• Sales pitches and slide decks
• Listicles

Detection

Count bolded phrases per paragraph. More than 3-4 bold phrases in a short paragraph is suspicious.

Inline-Header Vertical Lists

The Pattern

AI outputs often include vertical lists with this specific format:

• Number/bullet followed by
• Boldfaced inline header followed by
• Colon and descriptive text

Examples

Numbered variant:

1. **Historical Context Post-WWII Era**: The world was rapidly changing...
2. **Nuclear Arms Race**: Following the U.S. atomic bombings...
3. **Key Figures Edward Teller**: A Hungarian physicist who advocated...

Bulleted variant:

- **Concept and Lyrics** — The artist defines the theme and lyrics of the song.
- **AI Melodic Drafts** — AI produces different melodies following the prompt.
- **Human Supervision and Enhancement** — Producers adjust the instrumentation.

Broken List Markers

Instead of proper wikitext/Markdown, AI may output:

• Literal bullet characters (•)
• Hyphens (-)
• En dashes (–)
• Hash symbols (#)
• Emojis as list markers

This occurs when AI output is copied as plain text.

Detection

Look for:

1. Consistent **Header**: description pattern
2. Bullet points as literal characters instead of proper syntax
3. Every list item following identical structure

Emoji Usage

The Pattern

AI chatbots sometimes decorate content with emojis, especially:

• Section headings prefixed with emojis
• Bullet points marked with emojis
• Emphasis emojis inline

Examples

"Let's decode exactly what's happening here: 🧠 Cognitive Dissonance Pattern: You've proven authorship, demonstrated originality..."

"🪷 Traditional Sanskrit Name: Trikoṇamiti 🕰️ 1. Vedic Era (c. 1200 BCE – 500 BCE) 🔭 2. Sine of the Bow: Sanskrit Terminology"

Where It Appears

Most noticeable in:

• Talk page comments
• Draft articles
• Informal communications

Detection

Any emoji usage in formal article content is a strong AI signal. In talk pages, emoji-decorated headers or lists are suspicious.

Over-Use of Headers in Short Content

The Pattern

AI defaults to explicit section headers even when:

• Content is short (under 500 words)
• Prose would flow more naturally
• Headers fragment a coherent argument

Why This Happens

AI chatbots receive instructions like:

"Use a single main title (#) and clear primary subheadings (##)"

Example

A 300-word response with:

## Introduction
[2 sentences]
## Background
[3 sentences]
## Analysis
[3 sentences]
## Conclusion
[2 sentences]

Human Pattern

Humans typically use headers when genuinely helpful for navigation in longer content, not mechanically for every topic shift.

Excessive Parallelism in Lists

The Pattern

Every element in AI-generated lists follows identical grammatical structure.

Example

- Improving efficiency
- Reducing costs
- Enhancing quality
- Increasing satisfaction

All present participle ("-ing") form, identical structure.

Human Pattern

Some parallel structure with natural variation:

- Better efficiency
- Cost reductions
- Enhanced quality
- Higher user satisfaction

Curly Quotation Marks

The Pattern

ChatGPT and DeepSeek typically use curly quotes ("..." and '...') instead of straight quotes ("..." and '...').

Technical Details

Caveats

• Microsoft Word has "smart quotes" feature
• macOS/iOS auto-convert quotes
• Citation tools may preserve curly quotes from sources
• Some fonts display curly quotes as straight

Detection

Consistent curly quotes throughout text (especially mixed with other AI signals) suggests ChatGPT/DeepSeek use. Claude and Gemini typically don't use curly quotes.

Subject Lines in Messages

The Pattern

AI-generated messages sometimes begin with email-style subject lines:

Examples

"Subject: Request for Permission to Edit Wikipedia Article - 'Dog'"

"Subject: Edit Request for Wikipedia Entry"

"Subject: Request for Review and Clarification Regarding Draft Article"

Why This Happens

AI was likely asked to draft a message/email, and the user copy-pasted the entire output including the subject line.

Detection

Any text beginning with "Subject:" in a non-email context is a strong AI signal.

Over-Organized Structure

Pattern Summary

AI defaults to explicit organization even when unnecessary:

• Numbered steps for simple explanations
• Bullet points where prose flows better
• Section headers for short content
• Explicit "First... Second... Third..." sequencing

Human Pattern

• Organization emerges from content needs
• Prose allowed to flow without constant structuring
• Headers used purposefully, not mechanically

Detection Checklist

When analyzing formatting:

• Are all section headings in Title Case?
• Is there excessive boldface (>3 per paragraph)?
• Are there inline-header lists (**Bold**: text pattern)?
• Are there emojis in formal content?
• Are list markers literal characters (•, –) instead of proper syntax?
• Is the content over-organized with unnecessary headers?
• Are all list items perfectly parallel in structure?
• Are curly quotes used consistently?
• Does text begin with "Subject:" inappropriately?