.png)
.png)
A non-profit group has initiated a collaborative research initiative with academic, technological, and corporate partners to develop anti-corruption detecting technology.
Integrity Distributed (InDi) is building a platform to identify inappropriate payment trends in corporate third-party payment systems without requiring those corporations to give their raw payment data.
InDi consists of researchers from Massachusetts Institute of Technology (MIT) and Harvard Business School, technology startups Kona AI and Integra Ledger, five corporate law firms, and eight international Fortune 500 companies. The initial financing for the initiative came from the AB InBev Foundation.
Here is how it operates: The platform collects corporate third-party payment systems data in emerging markets, anonymizes the data, and then analyzes the data company by company to evaluate corruption risk across the supply side spending patterns of the business. Using hundreds of tests and behavioral algorithms, each business assigns a risk score to payments that have been identified in its own data. Each company's representatives or outside counsel analyze the deals with the greatest risk.
According to the project's white paper, a predictive model is then developed for each organization, "designed to proactively identify potentially improper payments based on the attributes of each transaction." Identifying the characteristics of problematic payments can be done without including the company's raw data.
Finally, the company's models are merged into a single "super-model" that permits all project participants to share insights while safeguarding data privacy and anonymity.
According to a white paper published by InDi in November, the predictive value of spotting a possibly unlawful payment is increased by 25 percent when companies interact as opposed to when each company's model is run in isolation.
The objective, according to Vincent Walden, CEO of Kona AI, is to discover significant characteristics of an improper payment, such as keywords or tests that they consistently hit on.
"The goal is to profile the attributes of a corrupt or fraudulent third-party payment," he explained.
PhD candidate at MIT who conducted anti-corruption research for InDi, Praneeth Vepakomma, stated that the project leverages on recent breakthroughs in a decentralized machine learning technique called split learning, which enables participating entities to train machine learning models without sharing raw data.
Vepakomma stated, "We integrate split learning technology from our MIT research with the know-how of anti-corruption experts and workflows to form a distributed model to detect vendor fraud, corruption, and circumvention of controls."
The collaboration enables corporations to "train machine learning models to identify potentially improper or corrupt payments," according to Bryan Judice, global head of compliance data analytics and monitoring at Panasonic North America, one of the eight companies involved in the initial stage of the research. "Participants benefit from the models’ ability to learn across all participant data sets without having to share their data with each other."
Jeannine Lemker, a member of the InDi board of directors and a former top compliance executive at Meta and Microsoft, asserts that countries have made significant progress in partnering to combat corruption. InDi utilizes technology so that businesses may do the same.
She stated, "There are immense possibilities for private companies to change the world together. InDi is a way to share collective wisdom on what corruption risk looks like and tackle it early."
Walden stated that for the platform's outcomes to continue to improve and become more robust, more organizations must contribute their data and analysis of that data to the initiative.
"As more companies are added to the cohort, it is our expectation that the super-model results will continue to improve, thus fueling InDi’s mission to help organizations fight global corruption," the organization noted.
By fLEXI tEAM