The effects of the pandemic continue to deeply affect every part of life. The fight is ongoing: businesses are trying to survive, adapt, and find new ways to operate under the current conditions. News outlets worldwide report rising Covid cases, stricter rules, and heavy lockdowns. Financial stocks are facing their toughest times since the 2008 crisis. During these vulnerable moments, scammers are quick to act and seek new methods to commit cyber crimes, as today’s situation provides plenty of opportunities.
According to Merchant Savvy report, the global losses caused by banking fraud has grown threefold since 2011 — from $9.84 billion to $32 billion in 2020. The very same research suggests that fraud will cost financial institutions more than $40 billion in 2027. In view of the magnitude of this threat, businesses must realize the necessity of being proactive and initiate prevention programs. Companies that put security at the core tend to reduce the cost of fraud attack response by nearly 42%.
The majority of organizations that rigorously investigate their cyberattacks are in a better place: they pour a lot of effort to strengthen control, rationalize operations and enhance staff morale. Machine learning and Artificial intelligence are extensively used by banks to provide secure services and maintain their good names.
I offer to analyze the challenges and weaknesses of Santander Bank, a global and well-established brand in the financial world. We will see how machine learning and leveraging data can help address those issues. But first things first: let’s start with the key pressure points that make things easier for cyber criminals.
Prerequisites for fraud
From a business perspective, financial institutions struggle to endure fast tracking of partners and third parties. During the onboarding process, there is a high possibility of a partner being not fully vetted and monitored which often leads to collaboration with infamous companies. The lockdown and closures of agents and banks makes it hard to deliver the due volumes in time. Now, the financial companies are under severe pressure of delivering new products to the market at unprecedented speed.
From a managerial standpoint, we are witnessing a resource shift. In other words, leaders of financial institutions are giving most of their attention to business models, operations, and how they suffer rather than fraud prevention. Therefore, budgets shrink as all the other activities are deemed non-essential. The lack of due focus, any fraud attack investigation might be frozen. As far as the employee aspect is concerned, the remote mode has deprived the staff of infrastructure and corporate security measures. Additionally, the massive employee reduction or pay reduction in an effort to reduce cost may become a trigger for fraud. Unfortunately, those who leave could undertake theft or copy sensitive information to get back at the company.
Santander: leaving a window open
Santander Bank, one of the fifth largest European institutions, which now operates in North and South America, and Southeast Asia. Foreign expansion distracted the bank’s executives from the security area and reduced the control over the existing branches. The unit in Belgium faced data leakage. The branch’s blog domain had a misconfiguration which allowed its files to be indexed. Crucial info.json files containing Cloudfront API keys were among them. By acquiring access to those keys, a hacker could use the bank’s information — images, videos, documents and beyond — to their own advantage. After the leakage was exposed, the bank put in place security protocols.
In August 2020, Security magazine issued an article disclosing a large-scale scam hitting Santander’s ATMs in the US. The criminals managed to find a bug in the bank’s software, allowing them to use fake debit cards to withdraw the money. Over 100 people were arrested across the country, and a number of machines were closed until the glitch was fixed. The bank initiated an investigation and cooperation with the authorities. Overall, due to the ATM scam the bank lost tens of thousands of dollars.
The implementation of machine learning can become the first line of defense for Santander Bank and enhance their performance across the globe.
Machine learning in banking: the benefits
According to Autonomous Next research, AI technologies will help the banks reduce costs by 22% by 2030. Face recognition will be one of the most widespread counter-forces to combat credit card fraud. Therefore, financial institutions across the globe are in demand of robust AI systems. Executives around the globe comprehend multiple opportunities ML and AI can provide.
Algorithms open doors to automation and enhanced productivity. Delegating daily routine tasks to ML, managers can focus on strategic areas and complex challenges. Leveraging big data allows for personalization and a higher level of customer services. We are talking about optimization of the client footprint: analytical tools can track the slightest fluctuations in customer behavior and enables creating custom-tailored experience for each individual.
ML leads to better risk management for the bank and its clients. Algorithms generate comprehensive reports and make predictions based on analysis of transaction history. An institution knows risks before a threat occurs, therefore, it helps to be well-prepared and proactive.
Santander Bank: what can be done
Santander partnered with Cloudera, the enterprise data platform company, to build their own big data platform for its UK branches. Given their experience, Santander required a global system for running global business, as granular and sporadic solutions are unlikely to provide due visibility and control. With ML solutions, Santander Bank will have a big holistic picture of client behavior. Therefore, any deviant behavior will raise red flags, so the bank can take immediate action.
Machine learning solutions provide a new level of accuracy when it comes to fraud detection. The algorithms can be used to eliminate manual data verification which is prone to errors. Also, the technology can significantly reduce false declines and identify uncommon patterns in real time. Ultimately, ML can address the issue of high load, which can cause online services going down. In such cases, the bank receives an influx of client complaints. Machine learning helps a bank in terms of maintenance by continuous testing.
ML and AI are empty abbreviations without knowledgeable usage. Technology is just one side of the coin, and the other one is the corporate security posture. Implementing new solutions demands extensive employee education, commitment to security policies, and making sure everyone realizes own responsibility.