Platforms · Software · Providers · Markets
Answers include their data source
FM Intelligence has launched FMIQ, an AI assistant that responds to questions about FX and CFD brokers with figures, their time periods and the datasets they come from. Introduced this week, it gives portal subscribers a way to query information held by the firm, rather than relying only on material available in published articles.
The difference is visible in website-traffic figures. Asked about broker visits in 2026, ChatGPT returned OANDA's 14.6 million visits for January, a figure published in a FinanceMagnates.com analysis in February. FMIQ gave OANDA 155 million visits across January to September, drawing on the full FX/CFD heat map.
The portal holds broader coverage
The portal launched in March. Its volume tool includes almost 200 brokers, while its traffic heat map covers 98.
Other information includes platform, fines and licence records. FM Intelligence has also published nearly 200 analyses on its DataLab portal. Sylwester Majewski, head of insights and reporting, said some subscribers had found the amount of information difficult to work through.
FMIQ searches seven datasets and can handle follow-up questions within the same exchange. FM Intelligence says it is read-only, runs queries live and sends responses through an independent review before labelling them "Verified". The Ask FMIQ page also offers sample prompts, including a comparison of volume leaders with country-level traffic.
Public models depend on published data
Ramzi Ahmad, director of intelligence at FM Intelligence, tested ChatGPT, Claude, DeepSeek, Gemini and Grok on broker traffic, trading volumes and Cyprus Securities and Exchange Commission (CySEC) licences. With web search disabled, the models had no broker volume or traffic figures. When search was enabled, their answers drew on published material.
For volumes, the responses put EC Markets and TMGM at about $2.11 trillion and about $2.03 trillion a month, respectively. Those were FM Intelligence's second-quarter averages, published in July; the models did not fill places three to five. Finance Magnates reported on the test, which Ahmad said showed that web access did not provide a single checked industry dataset.
The gap matters because broker data can sit behind a portal login, while public models may only retrieve figures already published. FM Intelligence counted at least 10 brokers and platform vendors connecting AI agents to live client accounts in the first half of 2026. Those agents can read account balances and positions that would otherwise be unavailable to a chatbot.
What this changes for operators
For teams evaluating FMIQ, the immediate checks are the datasets it can query, the periods represented in each answer and what the "Verified" label means in practice. Operators should also establish how the assistant's read-only access is enforced, and whether the returned data is suitable for internal reporting or only for exploration.
The account-connected agents described by FM Intelligence raise a separate systems question: which balances and positions an AI service can access, and how that access is controlled. The source does not specify FMIQ's integration requirements, operating cost or review method. Those details are worth confirming before making it part of a broker's data workflow.
