
AI can already analyse trading data, identify unusual behaviour, summarise exposure and generate reports in seconds. Tasks that once required a dealer to move between systems, export data and manually compare activity can increasingly be accelerated or automated.
That raises an obvious question for the dealing desk: if AI can do more of the work, how much of the role still needs a human?
The answer is more complicated than either “AI will replace dealers” or “AI is just another tool.”
AI is likely to remove a significant amount of manual work from dealing. Routine monitoring, repetitive analysis and reporting are obvious areas where automation can reduce the time between a question and an answer. A dealer may no longer need to spend an hour identifying which accounts contributed to an exposure or whether a client’s behaviour has changed. Increasingly, that information can be surfaced before the dealer begins the investigation.
But surfacing a pattern is not the same as understanding it.
If a system flags a client whose trading behaviour has changed, somebody still needs to determine what that change means. Has the client altered their strategy? Has volatility changed otherwise normal behaviour? Is exposure concentrated in a way that is not obvious from the aggregate position? Are execution conditions affecting the result?
The same applies to risk. AI can calculate exposure, monitor thresholds and flag activity outside predefined parameters. What it cannot determine independently is what the brokerage should be comfortable with in the first place. Risk appetite, liquidity relationships, commercial priorities and market conditions all influence where those parameters should sit and when an exception actually deserves action.
This is where the dealer’s role begins to change.
Instead of spending as much time producing analysis, dealers may spend more time interpreting and challenging it. The value shifts from finding the information to deciding whether that information is complete, relevant and reliable enough to act on.
That also means dealing teams themselves may change. If automation allows fewer people to monitor a larger client base and investigate exceptions faster, some responsibilities will inevitably become less important. At the same time, new responsibilities emerge around supervising automated systems, validating outputs, defining escalation rules and deciding where human approval is still required.
The question, then, is not whether AI will leave dealing unchanged. It will not.
The more useful question is which parts of the dealer’s role become more valuable as the routine work disappears?
The answer is likely to centre on context, judgement, exception handling and accountability. AI can operate within a framework, but experienced people still need to define that framework, recognise when it no longer fits reality and take responsibility for the decision that follows.
So, will AI make dealers obsolete?
Probably not. But it may make the traditional version of the role obsolete.
The future dealing desk is likely to involve fewer manual steps between a question and a decision, with dealers spending less time gathering information and more time evaluating what the system is telling them.
That does not reduce the importance of the dealer. It changes where that importance sits.
This question was explored further during The AI-Powered Brokerage: How AI Is Reshaping Trading and Operations, including where AI is already being applied across brokerage operations, where automation adds value and where human oversight still matters.
Watch the full discussion: https://bit.ly/46JznXY