Pivoting to the New Way of Intelligent Decision-Making:The Journey Toward Convenience

With the advent of large language models, organizations have recognized the critical role of AI in their business frameworks. According to a recent BCG study, 89% of executives have identified AI and generative AI as “top-three” technological priorities for 2024.

AI, encompassing a variety of tools and technologies, offers expedited updates to fraud detection models and enhances the capacity to shift strategies, thus favoring proactive precision over-reactive measures. Today, organizations can accurately forecast emerging trends and fraud threats by harnessing historical data, integrating it into sophisticated AI models, and utilizing predictive modeling.

However, predictive modeling alone can lead to inconsistent decision-making without the support of comprehensive data and collective intelligence. To address this issue, organizations are incorporating the concept of signals into their payment and decision management strategies, employing various intelligent data enrichment tools.

Signals provide succinct, meaningful intelligence within vast, complex data sets, offering sigmoidal insights (i.e., probability between 0-1, distinguishing between genuine, fraudulent, compliance-related, business policy, environmental, promotional, and other trends). Currently, banks, consumer financing providers, merchants, and utility companies face the challenge of ensuring that their payments originate from legitimate versus fraudulent sources or comply with policy.

A signal focuses on probability, revealing the true nature of a transaction. A signal is an insight that can presented as a probability to use within an AI feature selection, process, or rule-strategy, adding a layer of precision to the decision-making process. This evidence validates the nature of fraudulent transactions but also builds consumer profiles to elevate the thinking beyond fraud and risk management to leverage intelligence for enhanced consumer journeys.

Common trends include:

• Card testing: Increasingly perpetrated by AI bots rather than humans, this fraud method validates a stolen card number for fraudulent purchases
• Synthetic ID account creation: A fraudulent individual creates an account using real and fake information to commit fraud
• ATO: An individual gains unauthorized access to a legitimate account

Signals informed by a consortium of trends can verify if a card is involved in a bot attack and will decline transactions without damaging trust and reputation.

In the case of a flagged transaction, it involves not just one signal but a combination of several indicators:

• Multiple transactions within a short timeframe across various organizations
• Historical patterns of fraudulent activity in the region
• Shared intelligence from the network indicating similar fraud attempts
• Compromised data from a recent data breach

This comprehensive perspective enables the system to make a more informed decision. While a single signal may not be definitive, multiple signals provide a solid foundation for decision-making.

As AI technologies rapidly advance, using signals becomes crucial to help organizations adapt their strategies based on historical data, especially against AI-augmented fraud. 

In this AI-driven era, organizations must take the lead in establishing trust with every new account created. Consider two simultaneous scenarios for a consumer attempting to request a loan, pay a bill, or purchase with a merchant, referred to as Event A and Event B:

Event A: Credentials are flagged as having been involved in a card testing attack. Further analysis of historical data indicates that the supposed owner of the account does not exist within the payment’s consortium. As a result, the transaction is declined, and the account is flagged – a clear case of synthetic identity fraud.

Event B: Conversely, a consumer with verified card information wishes to request a service (loan, bill payment, or merchandise), and the process is seamless. Although the card information is clear of any bot attack association, other details—such as time on file, behavior, patterns, and digital identity—are confirmed with their device identity, all in real time.

In both scenarios, signals are pivotal. They assist in declining or flagging fraudulent activities but ensure a smooth process for legitimate transactions. A signal is a data enrichment derived from extensive data analysis across large datasets extending from ACI’s expertise that allows for more variety and selection during an AI feature selection process or a fraud strategy.

In the scenario of Event A, if the flagged transaction decision is enriched by the signals across the AI, it will establish the legitimacy of the transaction in real time. Meanwhile, with the signal indicator for the legitimate transaction, the organization can establish trust with its digital identity for a seamless flow.  

Signals are indispensable in differentiating legitimate transactions from fraudulent ones. They offer a sophisticated approach to fraud prevention, moving beyond mere historical data analysis to a dynamic and real-time decision-making process.

As the world undergoes a technological shift, harnessing AI and machine learning to prevent threats is essential. Democratizing data and making intelligence accessible is crucial for forming an intelligent shield around payments.

Using signals will accelerate as organizations find cost-effective ways to prioritize consumers. These signals also help understand behaviors and preferences, supporting targeted efforts towards consumer centricity.

Signals are central to ACI’s intelligent decision-making. Our patented AI models and network intelligence provide access to critical insights, establishing ACI as a strategic partner for global organizations. Signals are not only proof of flagging an account or a transaction as fraud but are also used to deliver good transactions for customers and prospective customers. With nearly 30 years in the AI ecosystem and a dedicated data science team focused on enhancing the customer experience, ACI remains ahead of the curve.