Artificial Intelligence Banking Digital Payments Featured News

Embedded Finance and AI: Path to Efficiency or Crack in Security?

Embedded Finance and AI: Path to Efficiency or Crack in Security?

Artificial intelligence (AI) offers the potential to streamline crucial processes in embedded finance, such as KYC and AML. However, the main challenge for experts seeking to leverage AI solutions is to ensure that increased efficiency does not come at the cost of compromised security or bias in decision-making.

Since artificial intelligence (AI) burst onto the scene in late 2022 with the launch of ChatGPT, the technology has found applications in virtually every industry. In the financial services sector, it is estimated that two-thirds of fintechs currently incorporate AI into their offerings in one form or another. And the market for AI in fintech is only expected to grow—to more than $60 billion by 2031, at a CAGR of 22.5%.

Recommended: Best 6 Fintech Startups Transforming Business In 2023

Until recently, the most common form of AI in financial institutions (FIs) has been the customer-facing chatbot assistant. But the rise of so-called “generative AI,” like ChatGPT, has increased the prospect of AI being incorporated into other aspects of the financial industry, such as fraud detection, risk assessment, and customer onboarding.

One of the chief benefits that AI can offer is increased efficiency through the automation of manual tasks — a ‘zero-ops’ approach — resulting in a boost in performance that can be applied throughout diverse areas of banking operations. According to Tadas Bakutis, Chief Technology Officer at ConnectPay, it is no surprise that FIs are turning to AI to streamline processes, however, there are several technological and ethical challenges facing those who pursue deeper implementation.

“Technologically, there are challenges in ensuring that the AI systems used in compliance processes like KYC onboarding are accurate, reliable, and secure. Facial recognition systems, for example, may struggle with accuracy if the quality of the images used for identification is poor. Similarly, document verification systems may have difficulty recognizing forged or altered documents.” Bakutis goes on, “Ethically speaking, there are concerns that AI could lead to bias and discrimination, especially if the data used to train the system is itself biased.”

AI is most effective, Bakutis explained, when it is given very precise and localized questions to answer. For this reason, it has been shown to be highly capable in the customer-facing role of chatbot assistant at banks like Wells Fargo. “That said, AI has a propensity to provide factually incorrect responses. AI is a great potential tool to be trained and used to do some particular specific tasks, but it cannot replace human experts in general. Any result of an AI query should be validated by an expert in that area or domain.”

He also outlined that the primary challenge remains to be able to explain the technology to regulators to address the concerns around the use of AI, especially in compliance processes, to pave the way for widespread adoption. This notion is shared among the pioneers of the technology as well.

Recently, Open AI CEO Sam Altman addressed the members of the United States Congress with an appeal to regulate artificial intelligence. Overall, the regulators on both sides of the Atlantic are taking a proactive approach, as not too long ago The European Banking Authority (EBA) shared guidelines on the use of AI in the banking sector, covering issues like data quality and algorithmic transparency.

“When it comes to AI, both private and public sectors want more clarity on how to enable the technology to evolve further yet create the right regulatory oversight that would preempt any possible harm that comes with its growth,” he commented, “and this mutual effort may be the key to leveraging the best AI has to offer while maintaining risks under control.”

ConnectPay is an all-in-one finance platform for online businesses, offering a wide range of payment solutions, including SEPA and SWIFT payments, IBAN multi-currency accounts, Banking-as-a-Service, Corporate cards, and merchant accounts. All processes are operated via a fraud prevention and compliance management ecosystem. A smooth onboarding process and customized client solutions enable businesses to utilize innovative payment solutions to meet the needs of their digital customers. ConnectPay holds an EMI license, issued by the Bank of Lithuania, and is a member of the monetary authority of the Eurozone.

[To share your insights with us, please write to sghosh@martechseries.com]

Related posts

With Close of Fourth PDP Wellbore Securitization, Raisa Energy’s Total PDP Wellbore Securitizations Now Exceed $1 Billion

Business Wire

Moody’s Analytics Wins Best Decision Management Solution at AI Breakthrough Awards

Fintech News Desk

Everyware® Honors Small Business Saturday To Help Local Businesses Thrive during the Pandemic

Fintech News Desk
1