The Big AI Engines Barely Agree on Sources — and That Changes How You Do AEO

Multiple 2026 studies keep rhyming. An Australian Intender analysis of tens of thousands of citations found the major engines largely disagree on sources. Other trackers watching ChatGPT, Claude, Gemini, and Perplexity report tiny URL-level overlap with Google’s classic top results and with each other.

Engine personalities (useful caricatures)

• Perplexity: cites widely, often leans on Reddit/YouTube-style sources.
• Gemini: long answers, broad nets.
• ChatGPT: fewer sources, more primary docs; sometimes the only one tagging clicks with measurable UTM patterns.
• Claude: structured, selective; sometimes declines to name local businesses at all.

AEO implication

One “AI SEO” checklist will not travel. You need engine-specific monitoring and content that can survive different citation appetites: primary sources for ChatGPT-like behaviour, community and multimedia proof for Perplexity-like behaviour, directories and review graph strength where Claude points users instead of naming you.

If your AI visibility report only samples one assistant, you are tasting one spoon of the soup.