Long reads

3 ways generative AI is driving innovation in financial services

Sandra O'Connell

Sandra O'Connell

Journalist, Sandra O'Connell

It took sliced bread five years to gain widespread adoption. It took ChatGPT (apparently the best thing since) just two months.

Launched to the public in November, by January ChatGPT had reached 100 million users, making it the fastest growing consumer application ever.

It’s not just ChatGPT either, Google’s Bard and image generating applications such as Midjourney and DALL-E, as well as code generating systems like Copilot are all fuelling public interest in generative AI.

Governments have rushed to respond, from the White House’s Blueprint for an AI Bill of Rights and the UK’s policy paper “Pro-innovation approach to AI regulation”. The EU and China were already on the case but likely moving faster.

Some in the financial services sector, spooked by security issues including data breaches, have curtailed the use of generative AI programmes (GAI) by staff, countries too. In April, Italy briefly banned ChatGPT.

Others have moved just as fast to capture gains. For financial services, already well versed in the benefits of chatbots to customer service, that’s no surprise. GAI boosts the prospects of conversational banking across a range of channels.

The government believes it can harness AI to turbocharge growth. Ensuring fairness in AI systems is critical but activity is currently subject to a variety of legal frameworks, including data protection, equality, and general consumer protection laws, as well financial services legislation, making it difficult for businesses to navigate.

Despite the risks, and the lack of clarity, some are forging ahead. Already there have been reports of Goldman Sachs piloting GAI to generate code. In a tight labour market with tech skills at a premium, expect more of that.

Personalising customer experience

A survey from chipmaker NVIDIA shows the top AI use cases in financial services are natural language processing and large language models (26%), recommender systems and next-best action (23%), portfolio optimisation (23%) and fraud detection (22%).

“Banks, trading firms and hedge funds are adopting these technologies to create personalised customer experiences,” it said, citing Deutsche Bank’s multi-year innovation partnership with NVIDIA to embed AI into financial services in a variety of ways, including intelligent avatars, speech AI, fraud detection and risk management.

Enhancing business operations

Computer vision and natural language processing are helping automate financial document analysis and claims processing too, says NVIDIA. AI helps prevent fraud by enhancing AML and KYC.

The biggest obstacle it found to exploring this brave new world of GAI powered banking is a decidedly analogue one––recruiting and retaining AI talent.

In some ways, GAI spells more of the same. According to the Bank of England’s research, machine learning, a sub-branch of AI, was last year already being used by 72% of respondents to a financial services survey.

While AML and fraud detection applications are widespread, ML was also used in banks for enhanced credit risk analytics and by other lenders, to automate loan underwriting.

Improving decision making

Even though AI already had a strong footing, GAI represents “a monumental leap” says Andreessen Horowitz, a VC. It enables banks to mine “troves of historical financial data” to quickly answer almost any financial question.

One of OpenAI’s pilots was with Morgan Stanley, which deploys GPT-4 to organise its vast knowledge base. Jeff McMillan, the bank’s head of analytics, data and innovation, likened it to “having our chief investment strategist, chief global economist, global equities strategist, and every other analyst around the globe on call for every advisor, every day.”

According to Bain Capital, GAI can help with customised marketing, developing audience specific messages to market new or existing products, increasing sales, and conversion. It can help with document processing and automating the process of “extracting information and forward-looking guidance from financial documents, such as from invoices and contracts.”

Despite the speedy start, full-scale change will take time. The sector is highly regulated and cautious by nature. It can’t afford the risk of hallucinatory bots providing rogue answers. Think of the risk of mis-selling suits. Slowly but surely, change is coming.

You do not have to be an AI expert to make a smart career move. There are great openings on the Finextra Job Board, like the three below.

Bloomberg has its own GPT product and the Applied Machine Learning Scientist - Data Technologies will work in the Bloomberg Engineering Data Technologies Department to find innovative solutions to some of the most interesting problems in the financial industry, using data science and machine learning.

BNY Mellon has an opportunity for a Global KYC Compliance and Advisory role which can be based in Poole or Manchester. The global financial services company is looking for someone to serve as a KYC subject-matter expert (SME), including, but not limited to: AML name screening, customer and enhanced due diligence (CDD and EDD), politically exposed person (PEP) risk, and specialised due diligence programs.

Monzo is on a mission to build the world’s best banking app and the Director of Data Science, Operations will be in charge of leading the team responsible for optimising operations and helping to support a great CX.

For more roles, using or developing AI, look to the Finextra Job Board and find the perfect opportunity that suits your skills and experience.

Comments: (0)