CiteAbility™ Labs

Open experiments in Citation Integrity™.

This is where CiteAbility™ tests and learns in public. The work here is early-stage and exploratory — a way to surface what the framework can do before anything becomes a certified product claim.

See the frameworkTalk to us

By CiteAbility™ · Last updated

Key takeaways.

  • Labs publishes exploratory work in progress, not finished product claims.
  • Experiments are run against real websites and real model outputs rather than simulated data.
  • The Citation Integrity™ Benchmark compares a Baseline AI Answer with an AI Answer with Citation Integrity™ Intervention from the same model.
  • Labs results may change or be withdrawn as methods improve.
  • CiteAbility™ cannot guarantee citation, recommendation, ranking or inclusion by any external AI system.

Current experiments

Work in progress and early findings.

Labs projects are deliberately unfinished. They let us share direction, invite feedback, and stay honest about what is proven versus what is still being explored.

In progress

Testing the Framework on SynaBot.ai

Results and write-ups will be shared here as the experiment moves forward.

In progress

Citation Integrity™ Benchmark

The same models answer the same questions as a Baseline AI Answer and as an AI Answer with Citation Integrity™ Intervention.

View experiment

Responsible claims

Labs work is not a guarantee.

We describe exactly what our work improves — measurable citation and evidence quality — and we are equally clear about what it does not promise.

Everything published under CiteAbility™ Labs is an experiment, not a finished assessment. Findings may change as the framework evolves, as models change, and as we test against more sources.

We do not claim that any Labs project eliminates AI hallucinations, guarantees truth, or replaces human review. We share it because transparency about uncertainty is part of building dependable AI.

Want to follow along or suggest a Labs experiment?

Tell us what you're building or publishing, and we'll come back with the most useful starting point.

Talk to us

FAQ

Frequently asked questions

The questions organisations ask us most often.

Are Labs results the same as product results?
No. Labs work is exploratory and may use draft methods. Product scores and certification follow the assessed framework and its published boundaries.
Can I take part in a Labs experiment?
Sometimes. Contact us with your website and sector and we will say whether current work is relevant to you.
Do Labs experiments prove AI providers use CiteAbility™?
No. They show what changes when evidence is presented differently to a model we control in a test. They are not evidence of adoption by any AI provider.