AI has reached a tipping point, both for personal and business use. We have a superficial alertness to its risks – among them, security, and fairness – but regulation is a nascent work in progress. To put it simply – we have not yet tamed the beast. The mass accessibility of platforms like ChatGPT may be a boon for students who are late on their essays, but can compliance teams safely leverage AI – or do the risks outweigh the rewards?

Integrating AI into Compliance

Introduced in November 2022, the data used to train ChatGPT was initially cut off in 2021. As a result, it was unable to verify whether the information it generated was factual, or provide references. Put bluntly – there was no certainty that it was always correct. Despite users now being able to access up-to-date information through ChatGPT, there remain concerns around its accuracy – as well as risks of bias or hallucination. Naturally, this rang alarm bells for banks and regulators who initially moved to ban the use of the Large Language Model.

Compliance specialist Global Relay has reported on the varied approaches financial regulators have taken to AI’s integration more broadly while legal regulations are still being developed.

AI – Risk or Reward?

Global Relay’s 2024 Compliant Communication survey of a compliance/risk/surveillance audience asked whether they think AI is a risk or a reward. People were undecided with 17% saying risk, 10% reward, 32% undecided, and 40% saying ‘Other’ – which broadly meant they thought it was too soon to say.

On their future plans, when asked whether they would be using AI in compliance workflows in the next 12 months, 43% said yes and 57% said no. The global split here was stark – 34% of North American firms said that they would use it within the year, compared to around 72% of both global and EMEA-based firms.

Beyond the geographical variations in risk appetite, AI is seen as both risk and reward:

“There’s no doubt that AI does offer reward, but it’s not instant or an overnight thing.”

  • Robert Nowacki, Technical Account Manager & Communication Surveillance SME, Global Relay.

Where to Start with AI

Bill Gates described AI as “the most important advance in technology since the graphical user interface.” But how can compliance teams leverage its power and benefits, while mitigating concerns around the technology – primarily privacy, fairness, and accuracy concerns?

For Global Relay’s Nowacki, it all starts with the data. He notes that financial data is often messy and unstructured, and for AI, that just won’t do. His conversations with financial firms have uncovered the main blocker to their implementation of AI – unstructured data from multiple platforms. Global Relay works with these companies to capture, store and structure it so that it’s ready to search, draw insights from, and align with AI technologies. He concludes: “If you want AI reward, invest the time into structuring your data first.”

A Work in Progress

AI has the power to generate more precise alerts to compliance breaches and reduce false positives – but only if used correctly. Martin Gaterell, Associate Director: Private Side Advisory with Monitoring & Surveillance, UniCredit GmbH, believes there are some societal blockers to mass adoption of AI in regulation. Firstly, employees are concerned their jobs will be lost to the automation it can deliver. AI must be accurate enough to reduce false positives, prove its ROI to senior executives, and be secure and private for clients and workers. Gaterell opines that once a firm goes “all in” on AI, securing a 90%+ reduction in false positives – then mass adoption will likely follow.

While AI has clearly proven its usefulness in efficient automation across a broad range of industries, its nascency in regulation and explainability means financial firms still must approach with caution. Until there is a focused, unified stance on risk – particularly in the US, where a new government may refine the current Executive Order of the Biden-Harris administration. As Chip Jones, Executive Vice President, Compliance, Global Relay, stated:

“The jury is still out in the U.S. regarding the efficacy of AI in financial services compliance. Before U.S. financial services firms fully embrace AI to assist with compliance, these firms will need to see definitive data that demonstrates that AI is infact assisting in reducing the compliance burden.”

For more on Global Relay, head to their website.



from UC Today https://ift.tt/wbz2rsQ