Will AI tackle the payments problem? Some say yes, some say no.
Avoid the moustachioed money launderers and look carefully for genuine AI in payment processing
One of the buzzwords of the past five years has been Artificial Intelligence, or for those who prefer two letter acronyms to infuse their jargon-speak, AI.
We live in a soundbyte-orientated world, and the AI acronym is being bandied around with equal verve and vigor as other such marketing-spun phrases such as the absurd “Three Tier Curb” nonsense uttered regularly by the British government which really means “do as you are told.”
Given the ease that has been demonstrated recently at which government organizations, banks and central issuers can grab power, do we really want robots taking over human tasks and then being programmed centrally?
According to some sources in the payments sector, a vitally important support industry for the electronic trading world, the answer is yes.
Payment processing and the ability to demonstrate genuine risk management and recourse to merchant services providers has been a moot point ever since the arrival of affiliate marketing orientated small and unregulated entities using MetaTrader back in 2004.
Visa and Mastercard quickly became very wary of the potential chargeback and fraudulent transaction risk created by retail novice traders using credit cards to send funds to untraceable offshore entities and recoiled in fear, giving rise to a motley selection of unscrupulous, donkey-jacket wearing Middle Eastern thugs who frequented the darkest corners of some of the retail affiliate marketing and FX conferences, hawking their money laundering services from under moustaches that look like earwigs.
All this served to do was alarm the authorities further, and now, with the NFA and CFTC in the US having banned the use of credit cards for funding trading accounts altogether, and bona fide firms finding it increasingly hard to find a suitable payment solution, AI is being used as the next marketing tactic for the payment industry.
Perhaps a faceless algorithm – if it is a genuine algorithm – is the answer.
Over the past few years, the digital payments market has exhibited steady growth.
To keep up with the increasing number of transactions, companies are continuously looking for ways to utilize new tools that would help ensure smooth and efficient processes. Currently at the heart of this lookout is the use of artificial intelligence and its use-cases in reducing the number of false positives in fraud and AML monitoring, and facial recognition-based payment verification.
Marius Galdikas, CEO at ConnectPay, has shared his insights on the current trend of harnessing the power of intelligent systems and their role in fraud prevention and streamlining transactions.
AI-enabled facial recognition
The pandemic gave precedence to AI-driven facial recognition solutions. A group of restaurants and retailers in California combined the need for stemming the spread of Covid-19, and the task of handling payments securely. The effort resulted in a face-powered payment confirmation system, or PopID. According to Marius Galdikas, the pay-by-face idea bears great potential as it requires a lot less engagement from the customer’s perspective, which adds to its appeal.
“Such AI-powered payments decrease the required effort from the customer to the bare minimum,” said Marius Galdikas. “Eliminating the extra steps in the process—taking out the card, entering the required PIN—is likely to improve perceived shopping experience, as customers can focus on a grab-and-go approach and save time. This leaves very little room for hassle, which, in fact, may lead to increased shopping cart values too.”
While conducting a digital payment transaction, users want one thing above all else – a smooth, glitch-free experience, as unexpected lags are too much of a disruption for the modern-day customer. VISA has already attempted to bridge any possible outages by introducing a Smarter Stand-in Processing (Smarter STIP), which leverages deep learning to analyze past transactions before generating decisions to approve or decline transactions on behalf of issuers. The prototype is set to be released in October, and the smart stand-in solution may push other players in the payment industry to also look for additional measures that could help limit the number of declined transactions.
“For merchants, a smooth payment process may be the single most important aspect in terms of retaining customers with the ever-decreasing attention span,” said M. Galdikas. “Bypassing issues related to system glitches could help avoid costly failures for both PSPs and merchants. In addition, combining such solutions with AI enables to adopt a more dynamic approach and deal with similar situations in a timely manner, without any noticeable mishaps for the customers.”
The past few months reemphasized the importance of anti-fraud measures, as having more users switching to online shopping instead of brick-and-mortar businesses resulted in skyrocketing levels of scams. The finance sector has already ramped up the cybersecurity spent to keep the fraudsters at bay.
AI can assist with recognizing patterns and exceptions, minimizing fraud for complex, high-volume transactions. Human error is one of the more pronounced weaknesses, so using task-specific AI to recognize dubious transactions will have a significant impact on the overall fraud resistance of digital payment systems. In addition, fraud prevention not only protects against the loss of funds but also saves businesses additional costs for legal settlements, which can add up to above $3 for each dollar lost to scammers.
“The circumstances surrounding Covid-19 and the growth of online fraud adds up to the stressors that urge both merchants and PSPs to deepen their search for novel security tools even more,” explained M. Galdikas, “thus AI-driven solutions are highly likely to become a must-have among tools for ensuring transparency and reducing fraud.”
Without a doubt, the real impact of AI usage in the digital payments market will reveal itself over time. That said, it seems a wider implementation of AI-driven integrations is inevitable, as it carries the promise of next-level actionable solutions that would sustain the growing demand for digital payments.
Certainly facial recognition is a method which compliance divisions of companies will likely find useful in order to keep their reporting records straight with the regulators.
Hong Kong based ATFX recently deployed facial recognition technology for onboarding clients. Speaking to ATFX’s senior management team at the time of release, the rationale behind the move toward biotechnology was explained to FinanceFeeds.
“ATFX attaches importance to the security of customer data, so we have used various technologies to prevent risks steadily. This technology will help enhance the security standards and precision in client’s identity verification at offices, thus reducing identity theft in account opening” said the company’s senior executive.
“The company has injected investment into the latest fintech solutions to provide one of the best user interfaces in the industry currently. Our clients will be able to enjoy the benefits from the existing and new innovative services that the fintech arm is offering.”
Yes, it is likely that the authorities will trust a genuine AI solution developed by a bona fide software provider, as long as it has proper integration with merchant services providers, and that the AI is real, rather than those aforementioned moustachioed thugs who use the word ‘algo’ to give the impression that their payment solution is somehow a proper one which uses technology to find an appropriate channel when really the opposite is true – it’s just a bank account in Georgia or Latvia, to which they move the money under a different description and then transfer it to your brokerage.
That is called money laundering and has been the target of many regulators.
Stick to the genuine firms with provenance as software technology companies which have genuine relationships with Visa, Mastercard and Tier 1 banks, and the genuine AI should be a boon.
Just avoid the thick accents, billiard-ball heads and moustaches.