News

SAS to Offer ‘Model Cards’ to Enhance AI Safety

SAS has stepped up its “trustworthy AI” products and services by offering “model cards,” described as “nutrition labels” for select large language models. They will serve stakeholders across the business by automatically evaluating LLMs for accuracy, fairness and performance decay. SAS also launched an AI Governance Advisory group to advise customers on governance in the context of their own businesses.