Quick Start

Check for these immediate red flags:

1. Excessive significance language ("pivotal moment", "testament to", "underscores its importance")
2. Generic superlatives without specifics ("revolutionary titan of industry" vs "invented the first train-coupling device")
3. Present participle phrases tacked onto sentences ("contributing to the broader landscape of...")
4. Over-attribution of trivial facts ("According to local media coverage...")

Workflow

Step 1: Scan for signature phrases Look for AI-typical words/phrases:

• stands/serves as, is a testament/reminder
• vital/significant/crucial/pivotal/key role/moment
• underscores/highlights its importance/significance
• reflects broader, symbolizing its ongoing/enduring/lasting
• contributing to the, setting the stage for
• marking/shaping the, represents/marks a shift

Step 2: Check for statistical regression

• Does text smooth over specific facts into generic statements?
• Are unique details replaced with bland, positive descriptions?
• Does it sound like it could apply to many similar topics?

Step 3: Evaluate emphasis patterns

• Undue emphasis on significance and legacy for mundane subjects
• Over-attribution of basic facts to sources
• Superficial analysis attached with "-ing" phrases
• Claims of "active social media presence" or similar generic statements

Step 4: Look for structural tells

• Lists of media coverage without content summary
• Entire sections asserting notability
• Hedging preambles followed by importance claims anyway
• Over-emphasis on ecosystem connections (for biology topics)

Examples

Example 1: Generic Significance Language Input: "The Statistical Institute of Catalonia was officially established in 1989, marking a pivotal moment in the evolution of regional statistics in Spain." Analysis: ✓ AI likely - uses "marking a pivotal moment" for routine government establishment

Example 2: Over-Attribution Input: "According to ABC News coverage, the restaurant underscores its role as a well-known late-night venue." Analysis: ✓ AI likely - attributes trivial observation to news source unnecessarily

Example 3: Superficial Analysis Input: "The population stood at 56,998 inhabitants, creating a lively community within its borders." Analysis: ✓ AI likely - adds meaningless analysis with present participle phrase

Example 4: Human Writing Input: "Smith invented a new coupling device in 1887 that reduced train derailments by 40%." Analysis: ✗ Likely human - specific fact, no inflated language

Best Practices

• Don't rely solely on AI detection tools - they have significant error rates
• Look for patterns, not isolated instances - one phrase doesn't confirm AI authorship
• Consider context - some legitimate writing may use these patterns
• Focus on the underlying issues - policy violations, not just stylistic quirks
• Cross-reference claims - AI often makes unsupported assertions about sources

Common Pitfalls

• False positives: Human press releases and marketing copy can have similar patterns
• Model variations: Different AI models have different "tells" - ChatGPT vs Gemini patterns differ
• Evolution: AI writing patterns change as models improve
• Over-confidence: Even experts misidentify ~10% of cases
• Treating symptoms as problems: The real issue is usually policy violations (original research, unverified claims), not just the writing style

Progress checklist for comprehensive analysis:

• Scanned for signature AI phrases
• Checked for statistical regression (generic replacing specific)
• Evaluated emphasis and significance claims
• Looked for over-attribution patterns
• Assessed superficial analysis markers
• Cross-referenced factual claims
• Considered alternative explanations