Opportunities and Challenges of Frontier Data Governance With Synthetic Data
Not on the map. Why?
Abstract (arXiv)
Synthetic data, or data generated by machine learning models, is increasingly emerging as a solution to the data access problem. However, its use introduces significant governance and accountability challenges, and potentially debases existing governance paradigms, such as compute and data governance. In this paper, we identify 3 key governance and accountability challenges that synthetic data poses - it can enable the increased emergence of malicious actors, spontaneous biases and value drift. We thus craft 3 technical mechanisms to address these specific challenges, finding applications for synthetic data towards adversarial training, bias mitigation and value reinforcement. These could not only counteract the risks of synthetic data, but serve as critical levers for governance of the frontier in the future.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 21 of its 30 references that are in the map. First posted on arXiv 21 Mar 2025.
Builds on
21 of 30 references.
References · 21
- 2020Language Models are Few-Shot Learnerscited by 32,419
- 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,255
- 2014Explaining and Harnessing Adversarial Examplescited by 7,364
- 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,415
- 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 6,013
- 2017Deep reinforcement learning from human preferencescited by 3,912
- 2015Domain-Adversarial Training of Neural Networkscited by 3,044
- 2019Fine-Tuning Language Models from Human Preferencescited by 2,150
- 2019A Survey on Bias and Fairness in Machine Learningcited by 1,917
- 2020RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Modelscited by 1,255
- 2020Language (Technology) is Power: A Critical Survey of "Bias" in NLPcited by 956
- 2019EDA: Easy Data Augmentation Techniques for Boosting Performance on Text Classification Taskscited by 807
9 more
- 2017Generative Adversarial Networks: An Overviewcited by 590
- 2017Certified Defenses for Data Poisoning Attackscited by 303
- 2016Learning from Simulated and Unsupervised Images through Adversarial Trainingcited by 288
- 2018Manipulating Machine Learning: Poisoning Attacks and Countermeasures for Regression Learningcited by 225
- 2022Will we run out of data? Limits of LLM scaling based on human-generated datacited by 195
- 2018FairGAN: Fairness-aware Generative Adversarial Networkscited by 164
- 2024Observational Scaling Laws and the Predictability of Language Model Performancecited by 130
- 2022Equalizing Credit Opportunity in Algorithms: Aligning Algorithmic Fairness Research with U.S. Fair Lending Regulationcited by 9
- 2024Towards Data Governance of Frontier AI Modelscited by 3
Built on by
Citing papers per year: 2025: 1.
1 paper.