Regulation in 2025 - Our take on what we've seen so far this year

Over the course of 2025, we’ve made it a priority to speak to existing clients, other regulated firms and regulators themselves to try and better understand the challenges facing compliance professionals in today’s financial markets.

Along with our annual research report, these discussions help us make informed decisions about the products we offer, while allowing us to better help our clients with their regulatory requirements. Speaking to compliance workers at conferences, our own hosted roundtables, and in regular catch ups has shed light on a number of key topics impacting compliance - here are some of the most common themes we see come up.

AI - How is it impacting market abuse surveillance?

At a recent roundtable, we posed a question to the roughly 20 regulated firms in attendance: Who currently uses AI as part of their surveillance strategy?

Their response? Not a single firm did.

Despite the rapid growth of AI over the last several years and its increasing use across industries, the response to our question was not as surprising as it may initially seem. The cost of developing, calibrating and maintaining an in-house AI compliance solution, and then integrating that technology with any existing surveillance solutions, is simply not cost effective, especially when considering smaller firms.

However, while most firms would naturally be reluctant to implement their own AI models owing to the drain on resources and specialist knowledge required, many firms may not be aware of the role AI already plays in their own third-party trade surveillance systems. Indeed, good quality surveillance systems will likely harness some form of AI to aid in dynamically suggesting parameter changes, alert risk scoring, sentiment analysis of news articles, and other such complex tasks.

Perhaps, then, the response of the 20 firms we asked says more about how firms understand the tools they currently have implemented, how vendors explain the technologies used in their products, or some combination of the two. The reality is that the use of integrated AI in dedicated third-party trade surveillance systems is growing more commonplace by the day, with further adoption by both vendors and regulated firms to continue increasing in the near future.

The regulatory viewpoint

The adoption of AI by the regulators’ themselves has been perhaps more significant than that of regulated firms.

Across the globe, we are starting to see regulators begin to establish AI teams to both investigate the use of AI by bad actors hoping to manipulate markets, and explore its potential benefits as a tool to be used in surveillance systems.

In the case of the FCA, for instance, we have seen a number of AI initiatives launched in recent years. Last year, the regulator held a number of TechSprints (The FCA Market Abuse Tech Sprint - MASTS) examining, among other things, the potential uses of AI as part of market abuse surveillance. eflow’s head of surveillance, Jonathan Dixon, contributed a presentation examining “how AI and Machine Learning (ML) can be used to drive an improvement in both analyst efficiency, by highlighting high-value alerts, and also examining how clients can have parameters that are dynamically adjusted”.

The FCA has also set up an AI-powered “Supercharged Sandbox”, giving firms the opportunity to experiment with AI not just in compliance, but across other business functions as well,

This and similar initiatives by global financial regulators indicate a clear enthusiasm around AI adoption. Despite the possibility of new risks being introduced by advances in this technology, the regulatory viewpoint of AI is, generally speaking, an optimistic one as long as the solutions remain repeatable and explainable; black box solutions will not escape regulatory ire.

Collaboration vs. enforcement - what is the role of the regulator?

When we consider the role of financial regulators, we often immediately think of hefty fines and sanctions levelled against firms found to have broken the rules. This, however, is a view many regulators are eager to change.

We are increasingly seeing a shift towards a more collaborative approach to regulation. In the U.S. FINRA is implementing a number of initiatives intended to assist member firms with the regulatory burdens. They are investing in tailored guidance, clearer feedback loops, and more efficient oversight practices, aiming to reduce friction while giving firms more room to innovate and grow. Echoing this view, at the recent FINRA Annual Conference, CEO Robert Cook stated that “member firm engagement is critical to what [they] do.”

In the UK, we’ve seen similar views espoused by the FCA. Their recent Five Year Strategy places a huge focus on enabling growth, educating regulated firms, and “smarter” regulatory enforcement. In a recent speech by Therese Chambers, Joint Executive Director of Enforcement and Market Oversight for the FCA, she said that the regulator’s enforcement should be “predictable, proportionate, and purposeful.”

Enforcement, while not quite a last resort, can be a necessary tool used by regulators to declaratively stamp out abusive trading while signalling their intentions to other regulated firms. Despite this, the priority of the FCA, FINRA, and other global regulators presently is to facilitate growth while working with firms to help them better understand and meet their regulatory obligations.

How important is it to invest in trade surveillance technology?

The question, “do I need a trade surveillance system” is, by this point, generally accepted as being moot. It has largely been accepted that, if you want to trade on capital markets, you will need some form of automated trade surveillance system in place.

However, what still causes some confusion amongst regulated firms is exactly what your surveillance system should be able to do and, by extension, how much you need to work with trade surveillance technology.

Increasingly, the regulatory view is that compliance is no longer a tick box exercise. Trade surveillance must meet two clear requirements:

• A surveillance system must be able to successfully monitor for market abuse in a manner tailored to a firms’ particular trading strategy and associated business risks
• Compliance teams must understand their surveillance systems’ calibration and be able to demonstrate and explain its effectiveness should a regulator request it

The importance of explainability has been a recurring theme in the FCA’s recent Market Watch updates. In Market Watch 79, the regulator emphasised the importance of stress testing, monitoring and regular audits to ensure the effectiveness and applicability of a firm’s market abuse surveillance function.

In order for a surveillance system to be effective, then, it needs to have the functionality required to monitor for abusive trading in a meaningful way, but it also needs to be properly managed by Subject Matter Experts, with regular testing to ensure its continued effectiveness.

What the future holds

These topics only scratch the surface of the conversations we’ve had over the past few months. Nonetheless, they tell us a great deal about the key challenges facing compliance teams in 2025 and beyond.

Anxiety around new technologies and how best to manage them seems to be a recurring theme. However, as AI continues to bed in and teams grow more comfortable in managing more complex surveillance solutions, the worries may subside and be replaced by greater regulatory confidence, fuelled by an increasingly collaborative regulatory landscape.

If you’d like to discuss your regulatory requirements in further detail, feel free to book a consultation or get in touch using a contact form.