Jon Light on AI for Brokers in the MENA Region

At iFX Dubai 2025, Jon Light, Head of OTC Products at Devexperts, shared insights on the future of AI-driven brokerage solutions in the MENA region. 

From trader support and localization to risk management and sensible AI adoption, he shared his perspective on where AI is heading and how MENA brokers should be thinking about it.

Impact on trader customer support

As a live dev company with over 800 developers, Devexperts has housed an internal AI and machine learning division long before it became the widely recognized term it is today. 

One of the key areas Devexperts has focused on is using advanced AI chatbots to improve broker-trader interactions and efficiency. The need for fast, round-the-clock support is particularly important in the MENA region.

“It’s a no-nonsense market in the MENA region. You need to respond quickly and make the experience authentic,” Jon explained. 

“With AI chatbots, brokers can analyze trader behavior and provide support 24/7, without needing someone behind the scenes at all times.”

When Devexperts implements chatbot solutions, they don’t operate in isolation. They pull in existing broker data to train the AI and make it an expert on the firm to maximize its capabilities. 

“We feed in data from broker websites, FAQs, past support tickets, CRM records—anything the broker wants to include. Once trained, our chatbots can handle around 70% of trader queries. 

They also understand trader accounts, their open positions, their past orders, so support teams can focus their attention on the complex cases.”

AI has now advanced past a tool for answering questions, with the ability to help traders make more profitable and risk-averse decisions while increasing engagement. 

Localization and multilingual support

The MENA region is one of the most linguistically diverse in the world, making localization both a challenge and an opportunity for brokers. 

AI-powered Natural Language Understanding (NLU) technology makes it possible to serve traders in their preferred language. 

“In the UAE alone, you’ve got people speaking almost every language you can think of. AI can detect a trader’s language automatically, and is currently working to recognize and adapt to dialects”. 

In addition to automated language detection and translation, AI allows brokers to tailor interactions to cultural preferences. 

Trader communication styles vary, some prefer formal language, others a more casual approach. Culturally appropriate communication is also important. AI can recognize these nuances and adjust its responses accordingly.

Risk management and B-book operations

Jon discussed three key areas where AI is having an impact. 

Making B-book risk management dynamic

Traditionally, brokers manually adjust their B-book settings, grouping traders and instruments into A-book and B-book categories.

“Historically, you might change your B-book setup a couple of times a day based on market conditions and trader activity. But AI can make this dynamic, adjusting orders just before/as they come in,” Jon explained.

 “AI also takes into account broker positions, trader positions, orders in the OMS, market volatility, and risk appetite, allowing for more efficient flow management.”

The outcome is that B-book risk management is no longer just a static model, adapting to market conditions and broader risk parameters. 

“Having this dynamic way of managing your B-book can really optimize the flow. This is one area that Devexperts is looking at.”

Flagging manipulation and bad actors

AI is also helping brokers identify bad actors. AI can spot these patterns early and alert brokers before they become a problem.

“AI can help flag traders trying to manipulate the platform, by exploiting execution delays, and engaging other problematic behaviors,” said Jon. 

Managing risk in prop and funded trader firms

Similarly, in challenges, AI will now help identify truly skilled traders using challenge accounts. 

“We see traders buying two separate challenges and placing opposite trades to rig the system, doing whatever they can to get through. With AI, we can analyze their trading behavior early on.

The best traders don’t dive into the most volatile markets. They tend to stick to a small number of instruments they understand well. AI can analyze those behaviors and help firms filter out traders who are just taking advantage of the system,” said Jon. 

With more data analysis on trader behavior, firms will be able to use AI to assess traders that are genuinely skilled and worth bringing into a funded trading program. 

Once traders move into a funded program, AI continues to monitor their risk profile.

“Some traders who shouldn’t have passed will go all-in on volatile markets, aiming for a big payout. AI tracks that behavior and feeds it back into risk management systems.”

Advice for smaller brokers in the MENA region adopting AI

While large brokers have the flexibility to experiment, smaller firms need to be more strategic about how they integrate AI.

“Big firms have big budgets, big teams. They can afford to make mistakes. If you’re a smaller broker, you need to be careful about what you invest in. A lot of what’s being marketed as AI isn’t always AI,” Jon cautioned. 

“Start with a real problem and find the right AI solution for it. Don’t create a problem just because you think you need AI.”

Smart AI adoption should always be about minimizing risk, reducing costs, and increasing revenue.

“A tool must be helping you achieve one of those three things. Really do your research”.

Bringing algo trading to more traders with AI

As the phrase goes, ‘the coding language of the future is natural language’. This will now start to have a wider impact on algo trading accessibility. 

“Before, if you wanted to build a trading algorithm, you needed coding skills. Now, you can just describe your risk appetite, the types of things you want to trade, your strategies, and AI can generate the script for you. That’s something we’re actively working on at Devexperts,” shared Jon.  

Next steps

Rather than a single big leap forward, we will keep refining current AI models.  

“We’re still early: chatbots, localization, all of it will be a continuous evolution. 

The most important thing brokers can do right now is collect data. We need to plan ahead to make sure we have the right frameworks and the right data available for when we need it”, concluded Jon. 

Whether it’s improving trader support, managing risk, or optimizing B-book strategies, the key is to identify and focus on AI solutions that solve real business challenges that your firm has. Don’t get lost in the hype and AI-coined tools flooding the market. 

Contact us

To learn more about how Devexperts can help integrate AI-driven solutions into your brokerage, get in touch with us today.