Summary
- MahiMarkets is targeting Dubai’s brokers and proprietary trading firms with agentic pricing and risk technology.
- The move aligns with DIFC’s ambition to become an AI-native financial centre.
- Financial-market AI adoption will be judged on control, supervision, risk management, and resilience, not only execution speed.
MahiMarkets is expanding its agentic pricing and risk technology in Dubai, targeting multi-asset brokers and proprietary trading firms as the emirate pursues AI-native finance.
The London-based pricing, execution, and risk technology provider says its engine uses specialised agents to act on pricing, spreads, and risk as market conditions move. Dubai is now a priority growth region for the company, which plans to expand headcount and deepen its engagement with the emirate’s financial services community.
The expansion follows Dubai International Financial Centre’s wider AI-native programme, which aims to embed artificial intelligence across legal frameworks, regulatory systems, talent development, infrastructure, and business operations. DIFC has said the initiative could generate $3.5 billion in economic benefits and create 25,000 jobs.
MahiMarkets is seeking growth in a region with broker density, active retail flow, commodities exposure, and expanding FX activity. Dubai’s regulatory environment also gives financial technology vendors a market that is trying to move faster than many legacy financial centres in London, Frankfurt, or New York.
Agentic systems in trading technology need careful framing. A pricing or risk agent is not the same as a general-purpose chatbot. Financial markets already rely on automated systems, from execution algorithms to pricing engines and risk controls. The new question is how far AI-enabled agents can respond to market conditions while remaining supervised, auditable, and aligned with firm risk appetite.
The appeal for brokers is clear. Pricing and spread management are operationally demanding, particularly across volatile markets and multiple asset classes. Teams monitoring screens manually can miss changes, react late, or apply blunt rules when liquidity shifts. Software agents that adjust within controlled parameters could help firms manage risk more consistently while freeing staff to focus on client relationships, strategy, and exceptions.
Markets punish automation that behaves unexpectedly. If agents are allowed too much discretion, or if their decision boundaries are poorly understood, firms can create new forms of model risk, operational risk, and conduct exposure. Supervisory controls, kill switches, testing, explainability, and post-trade review have to be built into the product rather than added later as governance decoration.
Dubai’s AI-native finance ambition raises those stakes. Embedding AI into the financial centre’s legal, regulatory, and operational fabric could attract firms that want to experiment with automated workflows and AI-supported compliance. It could also force regulators to define how AI agents are authorised, monitored, and held accountable when they influence financial activity.
The Gulf’s speed can be an advantage, particularly where legacy systems and old market structures do not constrain adoption as heavily. That same speed demands credible governance. Financial centres earn trust through supervision, resilience, and market integrity as much as through technology adoption.
MahiMarkets’ expansion is a useful signal of where financial AI may become commercially meaningful. Some of the most important deployments may not begin with consumer apps or office assistants. They may sit inside pricing, risk, compliance, and execution infrastructure, changing how firms manage decisions minute by minute. The firms that benefit will combine automation with tight controls, because in market infrastructure autonomy without accountability is leverage waiting for stress.
