AI RESEARCH
AI Security Research Should Better Incentivize Defense Research
arXiv CS.AI
•
ArXi:2605.23448v1 Announce Type: cross This work examines an imbalance in artificial intelligence (AI) security research: the field tends to produce work on attacking AI systems than on defending them. Drawing on related academic papers, we find biased attack-to-defense ratios across subfields, including federated learning, speech recognition, membership inference, large language models, etc.