- How wide is the threat landscape for financial fraud now, and what key types of attacks are testing the limits of financial institution protection systems?
- With AI and ML firmly entrenched as fraud-fighting tools, what are some of the weaknesses of current models using only historical or internal data?
- How can financial institutions use synthetic data to understand the performance of and strengthen their defenses against evolving fraud and money laundering schemes?
- What part must consistent communication and intelligence data sharing between institutions play to help prevent or limit financial fraud impacts in the marketplace?
- How can financial institutions best invest precious budgets and staff to create the most effective AI systems for fraud detection and protection, now and in the future?
There’s no doubt that financial crime threats are increasing daily, in both volume and cost to the industry and its customers. Amid the rising number of consumer and business fraud attack methods, the banking industry seems to always be playing a game of ‘catch-up’ with the criminals. AI have certainly helped to protect against scams and prevent money laundering, yet many of the best bank AI defences now in place are based on the use of historical, or internal sources alone.
Many institutions are simply falling behind as fraudsters constantly change their tactics and innovate new ways to wring ill-gotten gains from the financial system. Financial services companies must establish clearer, more open communications, and be prepared (and allowed) to confidentially exchange both internal and external intelligence data to bolster their own AI systems against transaction and account threats. Meanwhile, new regulations to protect customers led by the UK mandate compulsory reimbursements of scam victims with financial institutions splitting the responsibility equally between both sender and receiver institutions.
While the responsibilities and regulatory burdens grow quickly, banks, credit unions and financial services companies of all types can no longer rely on just their own internal structures and strategies. They must arm against new threats and malevolent actors with the latest technology, including synthetic data tools and intelligent, flexible AI systems and staff. Cross-industry cooperation and wider sharing of data are emerging as key pieces of the puzzle required for all parties to detect and deter attacks increasingly aimed at all participants - large and small - across the financial system.
Sign up for this Finextra webinar, hosted in association with Red Hat, to join our panel of industry experts as we discuss emerging ways to better design and use AI and ML models to protect against fraud.