AI and Quantum pack a dual punch for financial institutions

  0 Be the first to comment

AI and Quantum pack a dual punch for financial institutions

Contributed

This content is contributed or sourced from third parties but has been subject to Finextra editorial review.

Attorney and Certified Information Systems Security Professional (CISSP) Megan Moloney delivered the keynote address to the Northwest Association for Financial Professionals (NWAFP) annual Summit. The event brought together treasury practitioners, financial services and fintech providers, and community leaders to Bellevue, Washington, USA for a full day of sessions to discuss trending issues in the worlds of treasury, cash management, and investing.

The message from the global consulting firm Guidehouse’s Defense & Security Associate Director was insight-packed yet sobering: Artificial intelligence (AI) and Quantum computing aren’t just bringing society a new revolution; the two advanced technologies might represent the most compelling opportunities and chilling realities humankind has ever faced – at the same time.

Moloney, an attorney with a broad background in high level information security positions in government and business, including stints at the United Nations and FBI, offered hope, wonder, and many strong cautions as part of her theme of contrasting the past and future industrial revolutions experienced by humankind since the first in the 1760’s – with the rise of coal and steam powered engines and machinery – and continuing up to the now maturing digital revolution.

She took the audience through examples from the initial upheaval brought to the textile industry by ‘spinning jennys’ into the 1870’s (Second revolution – onset of electricity, gas, and oil), then the 1960’s, with the rise of computers and nuclear energy, and finally the early 2000’s when the internet, renewable energy, and digital technologies made their entrance.

The next industrial revolution is already here, right now, said Moloney, and the rise of AI and quantum computing power will be its hallmarks through around 2030 and beyond. Just like its predecessors from the 1700’s up to the beginning of the 21st century, this era will be marked initially by innovation, followed in order by disruption, uncertainty, regulation, and if everything works out, “trust” from society, she said.

AI and Quantum: Promise or peril?

The big question humanity faces now, emphasised Moloney, is “How can we steer away from the peril and towards the promise that both AI and Quantum bring?” Many elements of society and most areas of business are already experiencing and using various types of AI to run their businesses, and develop and manage processes for themselves, partners, and customers.

Quantum computing, or “quantum”, is defined by IBM as “specialised technology—including computer hardware and algorithms that take advantage of quantum mechanics—to solve complex problems that classical computers or supercomputers can’t solve, or can’t solve quickly enough.” That technology, no surprise to anyone, offers tremendous opportunity, and poses towering threats, to society that can’t be ignored, noted Moloney.

“Right now, we are we are in the AI moment”, she pointed out, moments later also sharing some of the benefits - and corresponding downsides of the latest digital revolution, including fast-increasing use of predictive and generative AI. “We have more access now to data than we've ever had before […] yet at the same time, there’s a lot of it that's irrelevant, or insubstantial or inaccurate. The masses sort of value the speed of information over expertise and accuracy. Similarly, that income gap continues to grow […] 81 of the world's billionaires hold more (wealth) than 50% of our population.”

FIs ask: Is faster always better?

How does all of this raging technical innovation impact financial services? Well, said Moloney, it represents good news, and potentially very, very bad news as well because the power of quantum computing is beyond anything we’ve ever known when it comes to data, which banking institutions use and store and share to process trillions of transactions in multiple variations every day. Citing a 53-qubit quantum computer owned by Google, she told the crowd its power is just a portent of what’s to come. “It (has) already performed a function that would take a cutting-edge supercomputer 10,000 years to do. It did it in 200 seconds.”

Faster processing is certainly a good thing, and it opens up many opportunities for efficiencies and of course broader operating capabilities and new products and services. The problem is, fraudsters and other nefarious parties, some sponsored by certain foreign nation states, are slavering to capitalise on this new technology just like everyone else, and they will, said Moloney.

“What these two technological advancements kind of have going is what I would call constructive interference, right? Because they're happening at the same time in the same space and they're happening in a way that elevates the other […] and so the impact is much greater. That's what's happening with AI and quantum: they are going to amplify all of those impacts. Everything that we have with AI is going to be amplified by quantum.”

AI usage is helping advance many areas of the banking industry

Most financial institutions of size, said Moloney, are already successfully using AI for fraud detection, even if these tools sometimes make mistakes or delay processing. “Yes, false positive rates are high, but they're getting better and they're doing a lot better than we could have done without them. For example,” she continued, “(The US) Treasury's AI powered fraud detection system recovered $375 million last fiscal year. That's a big deal. Similarly, MasterCard has deployed a generative AI tool that's capable of scanning a trillion data points. They've already said that they anticipate it to increase fraud detection by as little as 20% or as much as 300%.”

Banks need to take advantage, and that amplifying effect of quantum computing will help them do so to an extent only dreamed of with AI-only applications, said Moloney. “Whatever machine learning tools you have, whatever AI systems you have, whatever data you've pulled together is going to be something that quantum can use and multiply. So, the better your machine learning, the better your data.”

Identifying the major enhancements AI presently brings to financial institution operations as fraud detection, regulatory compliance, trend analysis, and customer experience, Moloney went on to compare these to the top five benefits identified across all of critical business and government infrastructures as part of a study done by the US Department of Homeland Security and sector risk management agencies, or SRMA. “It looks very much the same as those that you've pinned for the financial sector: Operational awareness, performance optimisation, automation of operations, event detection and forecasting.

Tempering her enthusiasm, Moloney urged close attention be paid to the new technology of this latest stage of industrial advancement. “We also know that with industrial revolutions, it's a two-sided coin, right?” continued Moloney. “I don't need to tell you guys that these things disrupted your industry, right? Mobile Banking chatbots on your side, reduced numbers of bank branches, cryptocurrency, blockchain, peer to peer lending, equity, crowdfunding, micro-investing. These are all things that have changed the way you fundamentally do business.”

Narrowing options to address “hairy” AI threats

Moloney next detailed some of the risks, placing them into categories of “To AI, From AI, and With AI.”

“When I talk about ‘to AI’, there are a lot of things that an adversary could do to mess with your AI because now you're getting these wonderful (and) competitive advantages and things like that from deploying AI in your organisations. Well, they also become a great target for attack. They can attack the actual algorithms, and change how they function. That's something that's very doable.”

Broader concerns of privacy, of monitoring and surveillance, and even of predictive control, can mean that “all of a sudden our choices are vastly narrowed”, Moloney said. “It might be more nefarious than that - when you think about potential state actors or with domestic surveillance and control - in terms of what they might do to AI to really attack us. Then there are attacks from Ai. These ones are a little bit more straight up. The things you would think about, right? An AI enabled cyber-attack is the most obvious. And we're seeing those all the time – and we're gonna see more.”

One of the most frightening and increasingly prevalent examples is how criminals and terror agents are using AI to help them detect vulnerabilities, often ‘watching and waiting’ patiently for an opportunity to strike, later. “That takes a lot of work off the hands of anybody that would wish to do (harm to) our systems, our nation, our sector. They don’t have to find those vulnerabilities, if they have an AI system that does that for them.” She also noted social engineering is getting much more sophisticated now, unfortunately, thanks to help from AI.

Yet another area of great concern is theft of intellectual property (IP), specifically “reverse engineering of IP” theft. “They are using AI to steal data while they're using it. They're stealing the data one way or another, whether it's with an AI enabled attack or not, from (many) organisations, primarily American targets. Then they use AI to reverse engineer it and they can get to that IP a lot faster.”

Weighing the costs, knowing the perils, containing the risks?

How bad is the overall threat? Moloney cited a congressional report from 2013, saying that IP theft costs had (conservatively) amounted to at least $300 billion per year in value stolen. “That was ten years ago, and that is ‘chump change’ compared to what's happening right now. And if you overlay that with the ability of AI to go ahead and reverse engineer all of that IP, it’s terrifying. It’s bad news for us because it inhibits innovation, because who's going to spend all the time and money to innovate something if it can get stolen and decoded by AI in less than a month?”

Moloney explained that much of AI is presently developed or built from open-source data. “It's not necessarily documented, and so you can't really see it, and you can't really explain it. You don't really know how to assure people that it can be trusted, that it is using good data in a good way. So, that lack of transparency and that lack of defensibility is a major concern.” However, she asserted, amplified bias is probably the “most significant” concern for society and financial institutions in particular.

“Because whatever you input into the systems, garbage in, garbage out. Amplified bias happens a lot. And so, when you use AI to scale something, but you have a limited or biased data source from the beginning, you're going to hit a problem there.”

AI can be scary, she says, but what about quantum and its own capabilities to amplify everything it touches? “It’s difficult to understand what that risk might be. It's uncharted territory, and it's impossible to put real numbers against it.” Moloney explained that a group (Quantum Alliance Initiative) had tried to measure surreptitious, seemingly low key cyberattack methods using quantum computers, and were gaining more adherents to the fight from across all areas of business and government.

Companies and regulators must collaborate to keep AI/Quantum on society’s good side

Speaking of the present and future of the financial industry, and re-touching on her initial five points along the revolutionary curve, Moloney addressed the last stage leading to trust of all this advanced technology - regulation - and what it can do for humanity, and against it if properly engaged. She asserted that questions of AI and Quantum computing power and their proper use must be embraced by financial services companies and their partners in supervisory agencies immediately, openly, and forcefully to ensure their capabilities are fully understood. That the vast capabilities of these new tools are aimed mostly, at least, toward doing good for society.

“However we move forward, we really need to communicate closely with regulators. And those regulators need to understand what we're doing, what our technology is, what our pain points are, what our concerns are in a very transparent way. We cannot ‘hide the ball’. If you hide the ball, the adversaries will get there before you do. You need to be collaborative.”

“This is an area where anyone who's touching risk needs to be concentrated heavily, and hard, and now. Because again, an attack could be undetected, can happen instantaneously, and it could be ubiquitous across the industry.”

Sharing a final example of the challenges and chilling possibilities of AI and Quantum Moloney noted that just a week prior to the conference, Warren Buffett had opined on the possibilities of artificial intelligence’s wide usage at his company’s annual investor conference. “AI”, said the 93-year-old ‘Oracle of Omaha’, “has enormous potential for good and enormous potential for harm. And I just don't know how that plays out yet.” Moloney’s conclusion: “If Warren Buffett is uncertain, it must be an industrial revolution.”

Comments: (0)

/ai Long Reads

Scott Hamilton

Scott Hamilton Contributing Editor at Finextra Research

Where is AI winning, and heading next in financial services?

/ai

Yi Ding

Yi Ding Assistant Professor at Gillmore Centre, Warwick

AI can make P2P lending safer and more fruitful for investors

/ai

John Barber

John Barber Vice President and Head at Infosys Finacle

Scaling AI in European banking: A practical framework for success

/ai

Hamish Monk

Hamish Monk Senior Reporter at Finextra

What is automation in financial services?

/ai

Hamish Monk

Hamish Monk Senior Reporter at Finextra

Understanding AI for personalisation

/ai

Editorial

This content has been selected, created and edited by the Finextra editorial team based upon its relevance and interest to our community.