AI Risk Management Platform for Insurance Industry

保险行业人工智能风险管理平台

• 批准号: 10067619
• 金额: $ 6.33万
• 依托单位: HOLISTIC AI LIMITED
• 依托单位国家: 英国
• 项目类别: Collaborative R&D
• 财政年份: 2023
• 资助国家: 英国
• 起止时间: 2023 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=10067619
• 关键词: AI; Risk; Management; Platform; Insurance

Holistic AI (HAI) will lead a consortium in the development of the first-to-market AI Risk Management Platform for insurance. The Platform has been co-conceived with our insurer and AI technology partners to enable insurers to evaluate AI risk and to price insurance; and to allow AI companies to manage and mitigate AI development and deployment risks via risk management tools and by accessing effective insurance. The consortium will help shape the rapidly emerging policy and regulatory landscape for trusted and responsible AI/ML.The consortium brings together HAI's deep expertise in AI, data and software development and risk assurance knowhow; academic and policy expertise at UCL; and end user stakeholders including insurance providers.The need for the HAI platform is driven by the rapid adoption of AI and ML across society. It is already widely used in sectors spanning finance and banking, automotive (autonomous vehicles), medicine and healthcare (robotic surgery, clinical trials, automated image diagnosis, etc.), and is increasingly being used in HR and legal decisions. With this growth and with increasingly complex use cases, there is increasing concern about the risks associated with AI, with legislation and regulations emerging rapidly across the globe, which are dramatically changing the landscape, and reflect (and place) risks upon companies utilising AI which they must protect against.HAI will lead a consortium to meet global needs of insuring against AI, ML and algorithmic issues. These can be catastrophic - evidenced by KnIght Capital's $450m bankruptcy due to algorithmic trading system issues. Key challenges for the consortium are to define which risks should be prioritised and measured, and to create the tools to enable this.HAI already provides a sophisticated suite of systems for AI audit and assurance, which break down algorithmic processes across multiple verticals, while considering use cases and user/operator risk and oversight. HAI currently works with enterprise companies like Unilever, evaluating their multiple AI projects for bias and failure, ready for multinational application in the emerging regulatory landscape.The new Risk Management Platform will focus on specific insurance Industry requirements, evaluating and quantifying potential risk within/from AI, ML and algorithmic applications. Leveraging our existing assessment capabilities with wider requirements it creates a pricing and assessment structure enabling the insurance of algorithms.This creates a distinct offering providing the first tools suitable for Insurer AI assessment, positioning the UK at the centre of a multi-billion dollar market

整体人工智能（HAI）将领导一个财团，开发首个面向市场的保险人工智能风险管理平台。该平台是与我们的保险公司和人工智能技术合作伙伴共同构想的，使保险公司能够评估人工智能风险并为保险定价；并允许人工智能公司通过风险管理工具和获得有效的保险来管理和减轻人工智能开发和部署风险。该联盟将帮助塑造可信赖和负责任的AI/ML快速崛起的政策和监管格局。该联盟汇集了HAI在人工智能、数据和软件开发以及风险保障方面的深厚专业知识；伦敦大学学院的学术和政策专家；以及包括保险公司在内的最终用户利益相关者。对AI平台的需求是由整个社会快速采用AI和ML驱动的。它已经广泛应用于金融和银行、汽车（自动驾驶汽车）、医药和医疗保健（机器人手术、临床试验、自动图像诊断等）等行业，并越来越多地用于人力资源和法律决策。随着这种增长和越来越复杂的用例，人们越来越关注与人工智能相关的风险，全球各地的立法和法规迅速涌现，这些法规正在极大地改变环境，并反映（并放置）使用人工智能的公司必须防范的风险。HAI将领导一个联盟，以满足针对人工智能、机器学习和算法问题的全球保险需求。这可能是灾难性的——KnIght Capital因算法交易系统问题导致4.5亿美元破产就是证明。该联盟面临的主要挑战是确定哪些风险应该被优先考虑和衡量，并创建工具来实现这一点。HAI已经为人工智能审计和保证提供了一套复杂的系统，该系统在考虑用例、用户/运营商风险和监督的同时，分解了多个垂直领域的算法流程。HAI目前与联合利华等企业公司合作，评估他们的多个人工智能项目的偏见和失败，为在新兴监管环境中的跨国应用做好准备。新的风险管理平台将专注于特定的保险业要求，评估和量化人工智能、机器学习和算法应用中的潜在风险。利用我们现有的评估能力和更广泛的需求，它创建了一个定价和评估结构，使算法的保险。这创造了一个独特的产品，提供了第一个适合保险公司人工智能评估的工具，将英国定位在数十亿美元市场的中心