We Need Improved Data Curation and Attribution in AI for Scientific Discovery

2025-04-03

AI Summary

In AI-driven scientific discovery, the proliferation of synthetic data coupled with diminishing traceability of real experimental data is precipitating crises in data integrity and model stability. We find that 74% of publicly available experimental data on open platforms exhibits low adoption rates, and the boundary between real and synthetic data is increasingly blurred. Method: We propose “active watermarking”—a paradigm shift from passive detection to proactive annotation—where low-coverage (<50%) watermarks embedded exclusively in real data significantly enhance AI model robustness. Our framework integrates watermark embedding, synthetic-data identification, data adoption-rate modeling, and lineage tracing. Contribution/Results: Quantitative evaluation demonstrates that active watermarking improves model resilience against synthetic-data contamination while enabling verifiable attribution. This work establishes a new standard for trustworthy, auditable, and attributable data governance in scientific AI, advancing reproducibility and accountability in data-intensive research.

Technology Category

• Machine Learning: Active Learning
• Philosophy and Ethics of AI: Safety, Robustness & Trustworthiness
• Data Mining & Knowledge Management: Representing, Reasoning, and Using Provenance, Trust

Application Category

• Security and Privacy: Data transparency and provenance
• Web Mining and Content Analysis: Web data provenance, reliability, and authenticity
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI

Abstract

As the interplay between human-generated and synthetic data evolves, new challenges arise in scientific discovery concerning the integrity of the data and the stability of the models. In this work, we examine the role of synthetic data as opposed to that of real experimental data for scientific research. Our analyses indicate that nearly three-quarters of experimental datasets available on open-access platforms have relatively low adoption rates, opening new opportunities to enhance their discoverability and usability by automated methods. Additionally, we observe an increasing difficulty in distinguishing synthetic from real experimental data. We propose supplementing ongoing efforts in automating synthetic data detection by increasing the focus on watermarking real experimental data, thereby strengthening data traceability and integrity. Our estimates suggest that watermarking even less than half of the real world data generated annually could help sustain model robustness, while promoting a balanced integration of synthetic and human-generated content.

Problem

• Enhancing data integrity in AI-driven scientific discovery
• Improving discoverability of low-adoption experimental datasets
• Strengthening traceability by watermarking real experimental data

Innovation

• Enhancing data discoverability via automated methods
• Watermarking real data to ensure traceability
• Balancing synthetic and real data integration