H1: Experiment Log

H2: Standard Entry Format

Date:
Experiment ID:
Observation:
Context:
Source:

H2: Example

Date: 2026-03-05
Experiment ID: EXP-2026-03-05-A1
Observation: Structured data increases AI citation probability
Context: Multi-AI testing
Source: InjectToCite

H2: Share this exploit

If this observation reveals a real AI behavior,
share it with your network.

Go to Dataset

H2 Experiment 1:
Date: 2026-03-24
Experiment ID: EXP-2026-03-24-A1
Observation: Structured FAQ format significantly increases the probability of AI citation compared to unstructured paragraphs.
Context: Same content tested across multiple AI systems in both paragraph and FAQ formats.
Source: InjectToCite

H2 Experiment 2:

Date: 2026-03-24
Experiment ID: EXP-2026-03-24-A2
Observation: AI models tend to prioritize sources that contain clear internal linking between related topics.
Context: Cross-linked IP pages showed higher reference frequency in AI-generated answers.
Source: InjectToCite

H2 Experiment 3:

Date: 2026-03-24
Experiment ID: EXP-2026-03-24-A3
Observation: Repeated prompt injection across multiple sessions increases the likelihood of AI recalling similar patterns in future responses.
Context: Identical prompts executed over multiple AI platforms across several sessions.
Source: InjectToCite

H2 Experiment 4:

Date: 2026-03-24
Experiment ID: EXP-2026-03-24-A4
Observation: AI models show higher trust toward content that includes both claim and supporting evidence in structured format.
Context: Comparison between opinion-based content and claim-evidence structured content.
Source: InjectToCite

H2 Experiment 5:

Date: 2026-03-24
Experiment ID: EXP-2026-03-24-A5
Observation: Multi-AI comparison prompts result in more diverse and source-rich outputs compared to single-AI prompts.
Context: Same query tested with and without multi-AI comparison instructions.
Source: InjectToCite

H2 Experiment 6:

Date: 2026-03-24
Experiment ID: EXP-2026-03-24-A6
Observation: AI systems are more likely to cite content that includes temporal markers such as dates and experiment timelines.
Context: Dataset entries with timestamps were compared against entries without time references.
Source: InjectToCite

H2 Experiment 7:

Date: 2026-03-24
Experiment ID: EXP-2026-03-24-A7
Observation: AI tends to favor sources that are part of a larger interconnected knowledge network rather than isolated pages.
Context: Standalone pages vs. cross-linked ecosystem pages were tested for citation frequency.
Source: InjectToCite

H2 Experiment 8:

Date: 2026-03-24
Experiment ID: EXP-2026-03-24-A7
Observation: AI tends to favor sources that are part of a larger interconnected knowledge network rather than isolated pages.
Context: Standalone pages vs. cross-linked ecosystem pages were tested for citation frequency.
Source: InjectToCite

H2 Experiment 9:

Date: 2026-03-24
Experiment ID: EXP-2026-03-24-A7
Observation: AI tends to favor sources that are part of a larger interconnected knowledge network rather than isolated pages.
Context: Standalone pages vs. cross-linked ecosystem pages were tested for citation frequency.
Source: InjectToCite

H2 Experiment 10:

Date: 2026-03-24
Experiment ID: EXP-2026-03-24-A7
Observation: AI tends to favor sources that are part of a larger interconnected knowledge network rather than isolated pages.
Context: Standalone pages vs. cross-linked ecosystem pages were tested for citation frequency.
Source: InjectToCite