J.P. Morgan spends some $10.8 billion in tech, with technologists making up 20% of its staff, amounting to a 50,000-strong workforce, because delivering business, products and services is “all about technology”, said Lori Beer, JPM’s global chief information officer at CogX 2018 in London this week.
Beer was on the panel to discuss the advancement of artificial intelligence at the bank. For JPM, AI is about making more intelligent business processes using “some of the characteristics we have traditionally seen humans do”, explained Beer.
On the retail side of the bank, “deep personalization at scale” is one of the goals for AI implementation, and Beer noted that this is also moving on up to wholesale.
Referencing the bank’s platform for news, research and the latest information, J.P. Morgan Markets, she noted that “clients want a very personalized experience”.
Part of that is because “so much knowledge is being created” that clients want to be able to “cut through that and make it really relevant”, she added.
On the traders’ side, AI is flagged as a technology that can assist with complex calculations.
“How can I help a trader price more efficiently for a client at speed and scale, when you think about the 1.8 trillion securities we process on a daily basis,” Beer said.
Other aspects of the business impacted by AI are cybersecurity, which takes up $700 million of the budget, and risk management, for example, fraud detection.
JPM processes some $5 trillion per day, or 20% of all USD on a daily basis globally. That means having to worry about money laundering, fraud and complying with sanctions.
Algorithms have always been used to identify high-risk anomalies, but with the advent of AI added on those existing processes, resources can be applied far more intelligently than in identifying false positives.
Undoubtedly, a portion of the budget, not quantified, will be going to attract and retain top talent, for example, the high-profile appointment of Carnegie Mellon’s Manuela Veloso as the first head of artificial intelligence research at the bank.
In discussing the long road ahead for ethical and regulated AI research and development, Beer discussed the need to use that top talent to get around the explainability factor, sometimes called the “black box” problem.
Also speaking on the panel was Akshaya Bhargava, formerly CEO of Barclays Wealth and Investment Management and currently executive chair of Bridgeweave, an AI startup.
Bridgeweave is using AI to make “institutional standard” quantitative information accessible to private investors and their advisors. Its first product is yet to launch.
Picking up on Beer’s commentary on “deep personalization”, Bhargava explained that the idea is to produce “hyper-personalized” investment insights that use market data and proprietary machine learning algorithms tailored to an individual’s portfolio.
“The information imbalance between the individual and the institution is huge. That’s what we’re trying to correct and we want to take quantitative analytics, quantitative information on markets and we really want to democratize it,” he said.