Marsh has made a measurable claim about its new AI placement platform: Broker WorkBench will cut London specialty placement times from the industry average of two to four weeks to days or hours. That is a more concrete commitment than any of its major rivals have publicly attached to their equivalent tools - and it is the claim that will determine whether Broker WorkBench is remembered as a genuine market shift or a well-marketed upgrade.
The platform, developed by Marsh Risk's UK Specialty team led by Dominic Samengo-Turner, manages placement requests across the London insurance market - matching risks with capacity, negotiating terms, and binding contracts for lead markets, digital followers, and pre-arranged follow-form capacity. All recommended placements continue to require approval from Marsh's brokers.
Broker WorkBench will ultimately be the route via which Marsh places all its UK specialty and wholesale business in London - a scale commitment that goes beyond a pilot.
Andrew George, president, specialty, Marsh Risk, said the platform was built with a clear purpose: "delivering better outcomes for our clients as they navigate an increasingly complex risk environment."
Broker WorkBench arrives in a London market where the race to digitise specialty placement is already well underway.
Aon launched Broker Copilot in June 2025, starting with US national property and London Global Broking Centre property teams before expanding to additional lines and geographies through 2026, according to Aon's own announcement - similarly using AI to accelerate placement workflows as part of Aon's $1 billion technology investment commitment. Artificial Labs has supplied digital broking and underwriting technology to BMS Group and others, and expanded its European operations in June 2026. Ki Insurance, backed by Alphabet's growth fund, pioneered algorithmic follow capacity in Lloyd's, enabling near-instant line-setting on qualifying risks. The Lloyd's Market Association published its AI Adoption Toolkit in April 2026, establishing governance standards that signal AI deployment in this market is now an expectation, not an experiment.
What Broker WorkBench adds is Marsh's own data and analytics capability at the centre of the process. The LMA's work on Core Data Requirements - standardised data formats embedded into placement processes over the past two years - is the infrastructure that makes platforms like this viable. Marsh's breadth of specialty placement data is the raw material that makes its version potentially more accurate than a newer entrant's.
The efficiency gains these platforms deliver are real. Two to four weeks of placement time reflects genuine operational friction: manual data entry, sequential negotiation, repeated transmission of the same information to multiple markets. A platform that standardises data, matches capacity algorithmically, and manages digital follow placement in parallel rather than sequentially can genuinely compress that timeline.
But the efficiency gain accrues primarily to brokers with the volume, data, and infrastructure to deploy it at scale. A Marsh placing hundreds of specialty risks per week generates the data network effects that make AI matching more accurate and the platform economics that justify the investment. An independent specialty broker placing dozens of risks per week faces a different calculation.
That does not make Broker WorkBench irrelevant to independent brokers. It makes it a competitive signal. The brokers who will feel the efficiency gap most acutely are those placing in the same London market specialty lines - marine, aviation, energy, construction, financial lines, cyber - where Marsh's platform is designed to operate. Where once the differentiator was relationship and expertise, it is increasingly also speed. The GlobalData 2026 UK Commercial Insurance Broker Survey found that 9.2% of UK commercial brokers now identify AI as their biggest threat - ahead of price comparison websites for the first time. Broker WorkBench is a data point that justifies that concern.
