Search for: All records

To combat the deluge of enterprise breaches, government agencies have developed and published a wealth of cybersecurity guidance for organizations. However, little research has studied this advice. In this paper, we conduct the first systematic analysis of government guidance for enterprise security. We curate a corpus of prominent guidance documents from 41 countries and analyze the availability of advice, the coverage provided by the advice, and the consistency of advice across countries. To facilitate detailed analysis and comparisons, we develop a tree-based taxonomy and quantitative comparison metric, and then apply these tools to analyze “essential” enterprise best practice documents from ten countries. Our results highlight a lack of consensus among the governments’ frameworks we analyzed—even among close allies—about what security measures to recommend and how to present guidance. 

Large Language Models are typically trained with next-turn rewards, limiting their ability to optimize for long-term interaction. As a result, they often respond passively to ambiguous or open-ended user requests, failing to help users reach their ultimate intents and leading to inefficient conversations. To address these limitations, we introduce COLLABLLM, a novel and general training framework that enhances multiturn human-LLM collaboration. Its key innovation is a collaborative simulation that estimates the long-term contribution of responses using Multiturn-aware Rewards. By reinforcement fine-tuning these rewards, COLLABLLM goes beyond responding to user requests, and actively uncovers user intent and offers insightful suggestions—a key step towards more humancentered AI. We also devise a multiturn interaction benchmark with three challenging tasks such as document creation. COLLABLLM significantly outperforms our baselines with averages of 18.5% higher task performance and 46.3% improved interactivity by LLM judges. Finally, we conduct a large user study with 201 judges, where COLLABLLM increases user satisfaction by 17.6% and reduces user spent time by 10.4%.