Agentic data science with the Data Agent Kit
cloud.google.com·6h ago
AI Research
arXiv CS.AI·25d ago·5 min read
Does AI Assistance Leave a Temporal Fingerprint? Detecting Overreliance in AI-Assisted Writing and Programming
TL;DR
AI assistance in writing and programming leaves a distinct temporal fingerprint, revealing overreliance on AI tools. This study identifies key differences in how AI contributions manifest across mediums and suggests a method for detecting academic dishonesty.
✦ Why It Matters
Implement temporal analysis tools to detect overreliance on AI in student submissions today.
Key Takeaways
- Analyze keystroke patterns to identify AI assistance in student work.
- Differentiate between collaboration and overreliance on AI tools.
- Utilize temporal features for academic integrity assessments.
- Monitor AI contributions in both writing and programming tasks.
How It Works
The study employs temporal analysis of keystroke data to identify patterns indicative of AI assistance. By comparing the timing and frequency of keystrokes, researchers can detect bursts of activity that suggest AI contributions, distinguishing them from typical human writing or coding behavior.