About Is It AI?
An AI writing risk review tool built to support fair academic integrity decisions.
What we do
Is It AI? (isitai.co.uk) is a UK-based AI writing detection tool that helps teachers, tutors, and students screen written work for AI risk. It combines AI-powered analysis using Claude with statistical pattern detection, and shows flagged passages with plain English explanations of why each was flagged — not just a score. The tool is privacy-first: student work is never stored and is processed in real-time only. Built by Paul Byrne, Is It AI? is GDPR compliant and designed for UK education.
For teachers, this means faster screening and better evidence for fair conversations about academic integrity. For students, it means understanding how their work might be perceived before submission.
We are based in the UK and designed for UK education, though the tool works for any written content in English.
What we do not do
- ✗
We do not provide proof of AI use
Our results are probabilistic signals, not definitive verdicts. No AI detector can prove with certainty that text was AI-generated.
- ✗
We do not store student work
Content is processed in real-time and immediately discarded. We never train on, share, or retain the text submitted for analysis.
- ✗
We do not replace professional judgment
Detection results should be one input into a broader assessment. Teachers know their students best.
- ✗
We do not claim 100% accuracy
False positives and false negatives happen. Different writing styles, ESL students, and heavily edited text can all affect results.
Privacy and data
Student data is never stored. When you submit text for analysis, it is processed by our detection models and immediately discarded. We do not keep copies of submitted content.
GDPR compliant. We comply with UK and EU data protection regulations. We only collect the minimum data necessary to provide the service.
No training on your data. We do not use submitted content to train our models or any third-party models.
SSL encrypted. All data is transmitted over encrypted connections.
Our fair use stance
We believe AI detection should support fair academic practices, not create a culture of surveillance and accusation.
For teachers: Use results as one signal alongside your professional judgment. Start with curiosity, not accusation. Consider the student's writing history and circumstances.
For students: Use the tool to understand how your writing might be perceived. If you use AI tools responsibly as part of your process, be prepared to explain how and why.
For schools: We recommend developing clear AI use policies rather than relying solely on detection. Detection is one part of a broader approach to academic integrity.
Common concerns
What about false positives?
False positives happen with all AI detectors. Some writing styles, ESL students, and heavily edited text can trigger false flags. Our multi-model approach reduces this, but does not eliminate it. This is why we always recommend using results as one signal among many.
Can students game the detector?
Some paraphrasing tools claim to evade detectors. Our multi-model approach catches many of these, but no detector catches everything. The best defence is combining detection with process-based assessment.
Is AI detection biased against certain groups?
Research shows that AI detectors can flag non-native English speakers more often. We acknowledge this limitation and recommend that teachers consider it when interpreting results, especially for ESL students.
Why does accuracy vary?
Detection accuracy depends on the type of text, the AI model used to generate it, and how much editing was done after generation. Shorter texts are harder to assess. We show confidence levels to help you understand the reliability of each result.
Why we built this
AI writing tools are changing education. Teachers need practical ways to screen work fairly. Students need to understand how their writing is perceived. Both need a tool that is honest about its capabilities and limitations.
I built Is It AI? because teachers kept asking me the same question: how do I know if a student used ChatGPT? The existing tools gave them a percentage and nothing else. I wanted to build something that actually explains what it found — flagged passages with reasons — so teachers have something concrete to work with.
Built by
Paul Byrne
Founder
20+ years in digital marketing and search strategy. Led search at MediaCom for brands including LEGO, Adidas, Shell, and Coca-Cola. Worked at Google. Ran client-side search at TripAdvisor/Viator.
Now running SearchIntel, helping brands navigate AI search visibility — and building tools like Is It AI? to support teachers dealing with AI in the classroom.
Connect on LinkedIn →Questions?
We are happy to discuss our approach, limitations, or how the tool might fit your institution.