AI Between Threats & Protection - Issue #439

Amnah Ajmal, EVP of Market Development for EEMEA at Mastercard, discusses transformative fintech trends shaping Eastern Europe, the Middle East, and Africa. In a conversation with Sally Mousa of Forbes Middle East at GITEX 2023, Ajmal delves into Mastercard's commitment to fostering financial inclusion and pioneering innovation in these burgeoning markets.

Banks are increasingly exploring blockchain technology for issuing digital bonds, aiming to streamline the traditionally slow and costly debt issuance process. The European Investment Bank leads the charge, having issued four digital bonds to date, including a £50mn two-year bond and an SKr1bn green digital bond. Digital bonds offer issuers cost savings, reduced intermediaries, and faster settlement times, with potential savings of 35-65% in issuance costs. However, the digital bond market remains nascent, with just $500mn issued in the year to September 12, compared to $6.3tn in US bonds alone, highlighting the need for further technological infrastructure and market confidence.

The Financial Times delves into why technology, particularly artificial intelligence, is increasingly preferred for verifying bank customer identities. Machines have proven to be faster and more accurate than humans in facial recognition, a crucial aspect of identity verification. The article highlights the efficiency of AI in analyzing images, its ability to learn and improve over time, and its superiority in handling complex verification tasks. The shift towards online verification has become mainstream, offering enhanced security and convenience in customer onboarding and ongoing account management. The piece also touches upon the potential of behavioral biometrics as an emerging field in identity verification.

Google has introduced DigiKavach, an early threat detection and warning system designed to proactively identify and analyze emerging financial fraud patterns. The initiative aims to understand the tactics of scammers, draw insights, detect threats, and collaborate with the wider ecosystem, including the Fintech Association for Consumer Empowerment (FACE) and the Cyber Crime Helpline (1930). Google contributes to DigiKavach through Gmail’s phishing protection, Google Play Protect, and Google Pay safety alerts, blocking over 99.9% of spam, phishing, and malware, scanning 125 billion installed Android apps daily, and sending safety alerts for suspicious transactions.

A PwC report highlights the growing concerns among business leaders in Ireland and globally regarding the potential use of generative AI in cyberattacks, anticipating "catastrophic" impacts in the next year. Despite these concerns, the report reveals a disparity in understanding the cybersecurity risks associated with generative AI, with only 45% of Irish respondents feeling knowledgeable, compared to 58% globally. Leonard McAuliffe, PwC Ireland Cyber Practice partner, emphasizes the need for agility and proactive security measures, advocating for a responsible approach to integrating generative AI in cybersecurity defenses.

Gary Gensler, Chair of the Securities and Exchange Commission, has voiced concerns over the financial risks posed by AI platforms used by financial institutions, stating that a financial crisis triggered by AI is “nearly unavoidable” within a decade without proper regulation. He highlights the risk of many institutions using the same AI base model or aggregator, potentially residing outside regulatory purview. This call for regulation is crucial, especially for startups in fintech and analytics, as it may lead to a reevaluation of AI usage descriptions and potential valuation adjustments.

Artificial intelligence (AI) is revolutionizing the financial sector, from risk management to investment banking. However, this transformation brings complexities in both financial services and regulation. The rapid growth of AI in finance presents regulatory challenges and opportunities that need attention. While AI offers potential solutions to complex regulatory problems, such as detecting fraudulent market manipulation, it also raises concerns like the "black-box" problem, where AI decision-making lacks transparency. Additionally, there's a risk of automation bias and potential biases in AI training data. To harness AI's potential while mitigating risks, a balanced approach, including "explainable" AI frameworks and global cooperation among financial regulators, is essential.