How does AIQ monitor AI narratives?
AIQ runs the user's prompts against the eight engines it currently tracks on a daily schedule, captures the full response, extracts and classifies the cited sources, scores sentiment, and tags themes. This builds a longitudinal record of how the story shifts over time and where the engines diverge from one another.
For each topic, AIQ runs the user’s prompts against the eight engines it currently tracks – ChatGPT, Gemini, Copilot, Perplexity, Claude, Grok, Google AI Overviews, and Google AI Mode – on a daily polling schedule. Every response goes through the same pipeline, so the results are directly comparable across engines and across time.
What the pipeline does, per topic, per day
- Poll the engines: the prompts are run against the eight engines AIQ currently tracks.
- Capture the full response: each engine’s complete answer is recorded, not just a summary or a score.
- Extract and classify sources: the sources each engine cited are pulled out and categorized.
- Score sentiment: each response is scored for how positively or negatively it frames the topic.
- Tag themes: the recurring narrative themes in the response are tagged.
The output: a longitudinal, cross-engine record
The result is a record over time: what each engine said today, how the story has shifted week over week, which sources are gaining weight, and where the engines diverge from each other. That divergence is the point. The same prompt about the same brand can return materially different answers across the eight engines AIQ tracks, and AIQ exists to surface those differences.
Last reviewed: 19/05/2026
