AI has been talked about since the very early days of computing and has attained mainstream use in recent years with the likes of Amazon’s Alexa and Apple’s Siri.

“Just as in the last 40 years, computation has enabled us to change the way we do business and create new products, AI will help us to make better decisions,” Carlos Kuchovsky, chief of technology and R&D at BBVA, tells Finextra.

“We are now looking at the ways in which it can help us change the way we operate and bring value.”

The Bank of England has recently reported that machine learning tools are in use at two thirds of UK financial firms, with the average company using it two business areas, which is expected to double in the next three years.

It may be through interoperation with cloud and blockchain technology that AI's capabilities will be fully harnessed.

AI

Utilisation of machine learning and artificial intelligence has become commonplace in everyday life, whether it be in search engines, music streaming services or internet shopping. This has also extended to financial services with its use found in intelligent chatbots for example. Behind the scenes, AI has also been harnessed to help firms make better investment decisions or mitigate risk in credit worthiness or in cybersecurity threats and fraud detection.

While many use cases of machine learning are not particularly new, the explosion in the amount of data available to train models has supercharged the possibilities of AI.

An example is the Discover Weekly playlist on Spotify, where music fans are given a personalised mixed tape of songs they might like every Monday. This popular service is built on the data gathered from the listening habits of the 217 million monthly active users that Spotify attracts.

Financial services are certainly rich enough in data on customer habits, but historically they have been poor at using it.

“Financial services need to build a solid data foundation by bringing all these separate data streams together, and once this is established endless opportunities will open up,” says Nicholas Merriman, chief technology officer of financial services at digital services provider, Avanade.

“For example, AI and machine learning, combined with the power of the cloud, could be used to ’robo-advise’ through validated models on topics like anti-money laundering, delivering further benefits to organisations and their customers alike.”

Blockchain

There will be concern over the reliability of the data employed in AI models, which would make it more difficult to set at ease the minds of financial services firms, never mind those of the regulators.

This is where blockchain technology can assist in demonstrating the reliability and provenance of datasets and preventing its misuse for nefarious or illegal activity by cybercriminals or even adversarial nation states and terrorists.

“Blockchain technology allows a chain of transactions to be stored, with each link representing a piece of history in the story of that asset,” says Leo Attwood, director at LexisNexis Risk Solutions.

“When the distributed element is introduced, where the chain is stored in multiple locations, it creates, in theory, a transparent history of transactions that is immutable and tamper proof.”

In areas like fraud prevention and data gathering, blockchain technology can offer “indelible transparency,” as Attwood describes it, and could be augmented by AI in financial crime mitigation for example. An overseeing system which can spot patterns pointing towards suspicious activities, by overlaying an AI algorithm over the ledger, would make the underlying blockchain technology even securer.

Cloud

Cloud is "the fuel for innovation" according to Jacqui Morcombe of banking solutions provider, nCino.

"It has many tangible benefits including speed, scalability and reduced IT costs."

The data involved in the above use cases for AI places a huge burden on firms regarding storage. Cloud computing enables businesses to reduce the overheads of infrastructure and increases the capacity for storage of the data required to power such innovations.

“Cloud technologies offer an agility that traditional co-location and on premises capacity management cannot compete with,” says Attwood.

“Another benefit is the ability to access and manage multiple storage locations with ease, introducing redundancy and data distribution to a level once only attainable by the military, government and blue-chip multinationals.”

Cloud providers like Microsoft have machine learning and AI services available in their clouds, which start-ups and fintechs will use to power their innovations. The amount of computing power needed to power intelligent decision-making or fraud analytics made such innovations hithetro the preserve of the major incumbents.

“Cloud computing has levelled the playing field,” Merriman adds.

“It enables financial service organisations, big and small, to offer more powerful and customer-centric solutions.”

Collective use cases

“AI, blockchain, cloud and quantum computing technologies each have the capability to create value by increasing efficiency, compressing workflows, decreasing cost and improving customer experience, but combining them amplifies their impact,” says Mitesh Soni, chief evangelist at Finastra.

“Multiple use cases come to mind across all areas of banking: high frequency trading, enterprise-wide risk, credit modelling, and end-to-end settlement.”

All these areas presently rely on the coordination of banks and other financial firms on both sides of any transaction, guaranteeing for example the buyer’s ability to pay and the seller’s ability to ship and so on.

Where such processes are often largely paper-based, a huge amount of manpower is involved and can prove a burden on time and resources.

AI can verify the accuracy of documents provided and find discrepancies far quicker than a human brain can, while blockchain enables every step of behaviour to be tracked, allowing all parties to see the exact same thing and at the same time.

Paula da Silva, head of transaction services at SEB, says: “With this transparency, payments can be settled far quicker, meaning sellers aren’t sitting around waiting for the finance needed to support subsequent transactions.

“This greatly improves cash flow would of course benefit a business’ profitability and customer friendliness.”

And beyond

AI, cloud and blockchain all add value to the financial services landscape individually, but combined they can, according to Noam Zeigerson of Tandem, “help achieve greater value and allow for understanding of any data a company holds.”

“These technologies are essentially waves of disruption that run parallel to each other and may even feed into each other,” Zeigerson adds.

While cloud provides the data storage requirements needed to feed AI models, AI expands the complexities of decision-making in smart contracts harnessing blockchain technology.

Therefore, it will be advisable to view these areas in a holistic rather than siloed sense as they become increasingly integrated in the years ahead and are driven further by other emergent technologies such as quantum.

Quantum computing’s processing power could enhance the problem-solving capabilities of AI and also offer stronger encryption techniques to shore up any chinks in the blockchain.

“Quantum is still very much in the lab,” says Soni, “but holds the promise to compress compute cycle time from years and seconds.”

AI, cloud, blockchain and quantum will be discussed at Money20/20 USA in Las Vegas. Check back here on 27-30 October for coverage from the event.

Jamie Crawley, Reporter, Finextra