For years, tech firms promised insurance professionals “insight” from their systems, only to leave brokers and underwriters wrestling with static reports and under-used dashboards. The arrival of AI tools that can understand spoken questions and respond in real time is starting to change that dynamic.
Instead of navigating menus or pre-built reports, brokers can now ask, out loud, for the numbers they need: premium movements, claims trends, line-by-line exposure, by client, class, or portfolio. Platforms like iBAIS 3.0 are among the first to bring this kind of capability into mainstream insurance system workflows, but the bigger story is what this means for how brokers and underwriters actually work.
For Angela Driessen-Clancy (main picture, left), director at Driessen Insurance Brokers, the shift has been immediate in day-to-day servicing.
“The most significant change has been in the speed and efficiency of data retrieval and report generation,” said Driessen-Clancy. Tasks such as checking a client’s claims history and what premiums they paid last year compared to this year can be arranged into a simple chart in seconds.
That speed is not just an internal efficiency gain. It is beginning to reshape the client conversation itself.
“This time saved has allowed us to concentrate on other things,” said Driessen-Clancy.
The most powerful use cases emerging from speech-enabled AI are happening in the room – during renewal and pitch meetings where decisions are being made in real time.
Driessen-Clancy recalled a recent renewal with a long-standing client who wanted to know whether premium increases over five years were in step with claims.
“They asked how their premium adjustments over the past five years compared to the claims they had lodged in the same period,” she said.
Before her firm’s adoption of advanced AI with speech capability, this question would have called for a reasonable amount of analysis and prep work in advance or a follow up after the meeting with a detailed report. The iBAIS technology she is using allowed her to query the system in real time and generate a clear side by side view of premiums versus claims history immediately.
“This immediate insight allowed us to demonstrate that their premium trajectory was consistent with their actual risk profile, and it opened the door to a proactive conversation about risk management strategies for the year ahead,” said Driessen-Clancy.
This ability to provide that level of transparency on the spot can reassure clients and reinforce the brokers’ role as a trusted advisor.
“What would once have been a delayed follow up became a decisive moment in the renewal meeting itself,” she said.
That example captures the core opportunity: speech-driven AI doesn’t just speed up reporting, it elevates the broker’s role in the room – from messenger of historical data to real-time interpreter of risk.
Underwriting teams are seeing similar benefits when they can pull live portfolio numbers on demand. Amanda Dixon (main picture, right), operations manager at a large underwriting agency, pointed to the way conversational queries have replaced manual spreadsheet work.
“The new AI feature has been fabulous at gaining insights into information that previously we were struggling to put together manually using multiple spreadsheets,” said Dixon.
In her case, being able to answer on-the-fly questions about year-to-date premium for a particular insurer, with prior-year comparisons, turned what would have been a rushed or partial response into a confident, data-backed discussion.
If brokers and underwriters can now talk to their data, the next challenge is learning to “speak AI” well enough to get reliable answers – and to know when not to trust them.
For Driessen-Clancy, the cultural shift has been as big as the technical one. “The biggest adjustment has been learning to trust the system’s speed and accuracy. For many years, we were conditioned to manually verify every figure, so adapting to instant AI outputs required a shift in mindset,” she said.
Another adjustment has been refining the way questions are framed. That idea – that the quality of the question determines the quality of the answer – is echoed on the underwriting side. Dixon notes that even powerful AI can mislead if the query is vague or conceptually wrong.
“The insights AI provides are invaluable, however, as with any AI tool, you need to understand how to formulate a question correctly,” she said.
For example, asking this type of AI what your firm’s total revenue is for 2025 will not give an accurate answer as it doesn’t factor in any third party pay aways. A more specific question asking for net profit will very likely generate a correct answer.
That discipline extends beyond wording. Early adopters are building a layer of governance around their AI use: validating key queries against existing reports, “pinning” trusted queries for others to reuse, and keeping human checks in the loop for any decision that materially affects clients or capacity.
There is also a growing awareness of data privacy and security. Many brokers have experimented with public AI tools to draft emails or summarise client information, sometimes without realising they may be exposing sensitive data. Systems designed specifically for insurance, with ringfenced models and automated stripping of personal identifiers, aim to keep those benefits without the leakage. But that only works if firms consciously move this activity into secure environments.
None of this erases healthy scepticism. Dixon admitted she still worries that some companies will lean too heavily on AI outputs without adequate checks. Yet her first live query – recreating a complex portfolio extract that had taken a colleague a full day – delivered the same result in seconds, and even picked up a human error. That mix of caution and proof is typical of where much of the market sits: intrigued, impressed, but aware that shortcuts in oversight could be costly.
For brokers and underwriters voice-driven AI is no longer a futuristic add-on. It is arriving inside core systems, reshaping how data is accessed and discussed. Those who invest early in the skills, governance and data quality to support it are likely to find themselves not just saving time, but having more meaningful, advisory conversations – with clients, carriers and their own teams.
