Would the computer have found him? Bernie Madoff, one of the worst investment scammers of all time
Is the financial advisor who advertises so eloquently on the Internet also trustworthy? Artificial intelligence developed by Frankfurt researchers is supposed to create clarity.
WIT specialists at Goethe University have developed an artificial intelligence that is supposed to help assess the seriousness of investment advisors. The software uses information from social media for this purpose. For their work, the researchers led by Peter Gomber assumed that trustworthy financial advisors deal with information in professional social networks as conscientiously or manipulatively as they do when working with their customers. They like to present themselves on platforms such as LinkedIn and Xing in order to maintain their contacts and win new business partners.
The Frankfurt researchers used self-learning algorithms to relate the self-portrayals of American investment advisors from the networks to information about detected fraud cases that were stored by the US authorities in publicly accessible databases. If the information provided by the consultants matched the official facts, the scientists interpreted this as a sign that the business people were credible – because the data from the authorities could hardly be falsified.
According to Gomber and his colleagues, the results are particularly useful for financial market controllers: They could use the study to develop their own monitoring programs and improve existing models.