The International Swaps and Derivatives Association (ISDA) has published a whitepaper examining how generative artificial intelligence can extract and digitise credit support annex (CSA) clauses. The study evaluated eight large language models (LLMs) across 60 CSA documents and found that when provided with CSA-specific information, the models achieved accuracy levels of over 90%. In addition, ISDA partnered with Arizona State University to develop a multi-agent framework for CSA extraction. The research focused on five key clause types: base currency, eligible currency, minimum transfer amount, threshold, and rounding provisions.