LLM Legitimacy Filters Block Synthetic Entities and Expose Burner SEO Tests

Feeding fabricated keywords into generative engines no longer forces them to hallucinate citations.

Google AI Overviews and ChatGPT actively run legitimacy tests that detect and block synthetic entity seeding, reveals Colette Nickel.

When prompted about nonsense terms planted for a ranking experiment, the models refused to process the inputs as real-world entities.

Instead of pulling the seeded data into AI answers, the systems explicitly unmasked the operation.

ChatGPT analyzed the query, flagged the synthetic keywords, and output a direct warning stating the prompt was a fabricated SEO test conducted by the practitioner.

The filter detects isolated manipulation attempts and completely strips their ability to poison the entity graph.