SBIR Phase I: Secure and Scalable Collaboration Platform for Effective Detection of Money Laundering and Fraudulent Transactions

这个小企业创新研究（SBIR）第一阶段项目的更广泛影响是通过缩小当前安全计算技术与现代反洗钱技术要求之间的差距来减少金融欺诈。银行和金融机构对创建安全的数据交换机制非常感兴趣。仅在2020年，全球反欺诈和反洗钱合规成本就超过1810亿美元。审计和交易监控占总支出的19%。在这两个类别中，一个个位数的百分比改善代表着价值数亿美元的市场机会。拟议的平台的一个主要好处是，有组织犯罪的复杂金融动态可以被瞄准和减少。除了反洗钱和欺诈检测之外，该提案产生的安全计算框架，即可扩展的快速安全数据交换平台，可应用于医疗保健，保险行业和国家安全等领域。这个小企业创新研究第一阶段项目旨在创建一个系统的解决方案，以克服目前在打击欺诈银行交易方面的局限性。有效发现欺诈和洗钱的主要障碍是缺乏对客户数据的全面了解。目前，每家银行关于客户金融活动的信息有限，无法从客户在其他银行的整体概况中获益。由于关键的数据安全和隐私问题，诸如创建中央数据交换实体之类的幼稚解决方案已经变得不切实际。在这个项目中，团队提出了一种新的方法，通过利用先进的安全多方计算（SMPC）来解决这一挑战。与隐藏客户数据特定部分的匿名化方法不同，安全计算协议即使在联合计算期间也能保证数据的机密性。在过去，由于巨大的计算和通信成本，SMPC一直是不切实际的。然而，该公司在理论和实践上都取得了突破，使SMPC在有效检测欺诈和洗钱方面更加可行。需要进一步的研究和开发，使解决方案真正适用于现实世界的反洗钱场景，从而产生一种可以大大提高欺诈检测的技术。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The broader impact of this Small Business Innovation Research (SBIR) Phase I project is to reduce financial fraud by closing the gap between current secure computation technologies and the technical requirements of modern anti-money laundering. Banks and financial institutions have a strong interest in creating a secure data exchange mechanism. In 2020 alone, the global fraud and anti-money laundering compliance costs exceeded $181 Billion. Audit and transaction monitoring accounted for 19% of this total expenditure. A single-digit percentage improvement in these two categories represents a market opportunity worth hundreds of millions of dollars. A major benefit of the proposed platform is that the complex financial dynamics of organized crime can be targeted and reduced. In addition to anti-money laundering and fraud detection, the secure computation framework generated by this proposal, i.e., a scalable and fast secure data exchange platform, can be applied to areas such as healthcare, the insurance industry, and national security.This Small Business Innovation Research Phase I project aims to create a systematic solution to overcome the current limitations in fighting fraudulent bank transactions. The main roadblock for effective detection of fraud and money laundering is the lack of a comprehensive view of client data. Currently, each bank has limited information about the client's financial activity and cannot benefit from a holistic view of the client’s profile in other banks. Naive solutions, such as creating a central data exchange entity, have been rendered impractical due to critical data security and privacy concerns. In this project, the team proposes a new methodology to address this challenge by leveraging advanced Secure Multiparty Computation (SMPC). Unlike anonymization approaches that hide a specific part of customers’ data, secure computation protocols guarantee data confidentiality even during a joint computation. In the past, SMPC has been impractical due to enormous computational and communication costs. However, the company has made both theoretical and practical breakthroughs that make SMPC more feasible for effective detection of fraud and money laundering. Further research and development is needed to make the solution truly applicable to real-world anti-money laundering scenarios, and thereby produce a technology that can drastically improve fraud detection.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.