Blog

How compliant traders can manage the generative AI synthetic data tsunami

Generative AI was last year’s technological innovation favourite, but its full potential for trading professionals is not yet entirely realised. Using AI for communications risk and compliance purposes in the sector should become commonplace argues Shaun Hurst, principal regulatory advisor at communications compliance firm, Smarsh.
By Shaun Hurst

Generative AI was last year’s technological innovation favourite, but its full potential for trading professionals is not yet entirely realised. Using AI for communications risk and compliance purposes in the sector should become commonplace argues Shaun Hurst, principal regulatory advisor at communications compliance firm, Smarsh. 

Last year saw a notable increase in the adoption of generative AI, such as OpenAI’s ChatGPT, Google’s Bard (now Gemini) and Microsoft’s Copilot. This trend represents an evolutionary leap in how companies across various sectors are integrating technology. As AI becomes a staple in the workplace, especially for traders and their complementary compliance teams, organisations are now faced with the challenge of understanding its full potential to identify and mitigate compliance risks, as well as manage the immense increase in the data produced.

In June 2023, McKinsey & Company outlined the potential value of these new technologies. It is predicted that generative AI could affect risk and compliance worldwide significantly, potentially to the tune of $250 billion. This is partly due to its ability to protect traders against bogus claims, identify gaps in communication that could indicate off-channel conversations and capture data in a way that maintains the context of all messages. 

While communications monitoring can be associated with micromanaging, there are significant benefits for traders, primarily, as it provides them with a line of defence if a client makes claims about alleged promises made. While of course, if a trader wants to communicate in a way that is not monitored, they can find a way to do so, firms should provide guidance and channels for compliant communication to mitigate against litigious action.

While our research indicates that 65% of highly regulated businesses are concerned about understanding new regulations, under a quarter see the lack of clarity on rules for generative AI as a major worry. This suggests that there is not enough urgency in grasping the innovative methods and processes that trading teams are using, as well as how to make the most of the synthetic data created by AI. 

Therefore, the challenge is to understand how generative AI can be used to assist traders in content management, risk navigation and regulatory compliance and then monitoring usage.

The potential of generative AI for trading professionals

As AI tools become more common, there is an increasing demand for systems that can handle AI-created content, especially in terms of compliance in highly regulated trading teams. The issue however, is twofold: firstly, companies must determine how to stick to regulations when they use generative AI. Secondly, they should be ready to use AI to help with compliance, especially as the technology generates so much data that managing risk manually is no longer realistic.

For instance, AI tools such as Microsoft’s Copilot can effectively double the content each employee creates, making it essential to have strong systems for saving and monitoring this data. Since regulatory requirements demand that trading teams collect and keep records of information, the systems used to archive this data must be upgraded to handle the huge growth in content developed by AI.

Monitoring the content AI creates and where it is then shared is something businesses are focusing on. Our research found that 71% of businesses in heavily regulated sectors are reconsidering their digital communication rules. This is partly because there have been more fines for communication that happens outside of official channels in many developed markets. 

While it is important to monitor for unfair trading practices in AI-generated content and the channels it is shared on, it is also crucial to navigate non-financial risk. The UK’s Financial Conduct Authority (FCA) is paying close attention to non-financial risks, including company culture and how employees behave. Advanced AI systems can help by quickly flagging potential red flags in the content of messages and the context they were shared in to detect bullying and sexual harassment.

Companies now have the ability to monitor content with AI, changing how they handle risks to their reputation. This is especially important for trading organisations that are increasingly worried about their public image and how their employees act. AI tools help keep an eye on activities that could damage their reputation, allowing companies to deal with problems early before they turn into bigger crises. 

Regulatory perspectives on generative AI

The FCA has been cautious to regulate using AI in order to promote growth in the technology sector and increase economic and business activity. However, as AI becomes more common in the industry, the FCA will need to update its advice on how to use it. We should watch what is happening in other countries to get an idea of the kind of changes that British regulators might think about in the future. 

The US Securities and Exchange Commission (SEC) has previously made statements about generative AI that are even more cautious than those of the UK’s regulator. SEC chair, Gary Gensler, has warned that AI could affect the stability of financial markets because it might lead to fraud and conflicts of interest. Additionally, president Joe Biden’s Executive Order on AI emphasises that regulators are more focused on protecting consumers than on promoting innovation.

Similarly, the European Union’s AI Act plans to sort AI systems into four levels of risk, based on how they are used. This way of looking at AI shows that compliance teams in the UK need to be prepared for stricter views on AI in trading in the near future.

Generative AI presents a tough test but also a great opportunity for trading institutions. Financial companies need to be flexible, using AI to handle the flood of data and keep their business running smoothly. If they do this, they can stay ahead with new technology, turning possible upheaval into a competitive edge.

The growing regulatory risk around AI

Firms can expect to add proof of AI usage to the ever-growing list of governance and reporting items that need to be maintained just in case a regulator or two come knocking, writes Firebrand Research founder and chief executive, Virginie O’Shea.

Operational resilience important in the context of T+1

As the US market prepares for an industry-wide transition to T+1 settlement, Stephane Ritz, managing principal, capital markets and T+1 global lead at Capco, discusses the need for organisations to reassess their operational strategy, resilience frameworks, and impact tolerance levels.

The multi-asset multiverse: Anticipating the trading landscape of the future

Following a volatile year for capital markets, Mark Govoni, chief executive of Liquidnet, delves into some key considerations for the ever-accelerating market, including: continuing multi-asset migration, the importance of emerging technologies in the pursuit of liquidity, and the evolving role of brokers.

Macro volatility will force market operators to reform

A range of macroeconomic, regulatory and market structure changes occurring across the world will drive the need for transformation for market operators across the globe, writes, Magnus Haglind, head of marketplace technology, Nasdaq, who points out that change doesn’t always come with a heavy price tag and substantial execution risk.

T+1 in the UK: Ignore the clock

As the US shift to T+1 settlement looms ever closer, John Bevil, senior solutions manager at Xceptor, discusses the potential for a similar adjustment on the other side of the Atlantic, the key hurdles for firms, and the importance of best practice irrespective of time constraints.

Charting the path forward as agreement on Mifid II revisions nears

After five years in practice, the European regulators are on the final stretch to agreeing changes to certain aspects of Mifid II. Anne Plested, MCSI, EU regulation, ION Markets, delves into the key milestones and implications of the imminent decisions.

Intraday FX swaps needs better tech for a prime efficiency revolution

As T+1 looms ever nearer, Alex Knight, head of EMEA at Baton Systems, delves into the effect of the shift on FX swaps, highlighting the opportunities they hold for market participants, as well as further establishing the importance of more sophisticated settlement technology.

Striking gold in hybrid solutions: Navigating the build vs buy debate

There is nothing new about the build vs buy debate – it has been going on for decades. There are pros and cons to both approaches. However, there’s also the hybrid approach, i.e. buying an off-the-shelf product and customising it, delivering, in many instances, the best of both worlds, writes Laurent de Barry, senior director, global head of solutions consulting at Exegy.

EU consolidated bond tape: Technology comes back into focus

In light of new developments around a potential consolidated tape provider (CTP) selection process expected to begin next year, Tim Whipman, head of business development at TransFICC, anticipates that modern interface technology, resiliency, and existing business continuity planning will be required when determining CTPs in each asset class.