Over the past few months, three announcements caught my attention.
Marsh Risk launched Risk Companion, bringing together proprietary risk data, actuarial models and broker expertise in an AI-powered decision platform. Aon has continued expanding capabilities such as Broker Copilot, Claims Copilot and its Digital Placement Exchange. WTW has elevated enterprise AI leadership while accelerating AI capabilities across the organization.
Three firms. Different approaches. One clear message: AI has become a strategic priority for the largest insurance brokerages in the world.
Most commentary naturally focuses on who is moving faster, which platform is better, or whether firms should build or buy.
I found myself thinking about something different.
I have very little doubt these investments can create real economic value.
Service teams should be able to handle larger books of business. Producers should spend more time selling and less time on administration. Brokers should reach markets faster. Claims professionals should resolve issues more efficiently. Institutional knowledge that once lived with a handful of experienced colleagues can become accessible across an organization.
That should mean lower servicing costs, higher producer productivity, greater operating leverage, faster execution and, for firms that implement AI well, meaningful EBITDA margin expansion.
Those gains are real. They matter. Boards should expect them.
But they are also unlikely to remain permanent sources of competitive advantage.
That is the part of the AI conversation I think deserves more attention.
What happens after everyone starts capturing the same operational gains?
History gives us a clue. Transformational technologies often create an advantage for early adopters. Then the technology diffuses. CRM, cloud computing, business intelligence and outsourcing all created meaningful operating advantages. Over time, they became part of the standard way companies operate.
AI may follow a similar path. Over time, the operational capabilities enabled by AI are likely to diffuse across the industry, just as previous waves of enterprise technology have.
Imagine the brokerage industry several years from now.
The major firms have copilots. Service workflows are increasingly automated. Producers have AI assistance. Claims teams use intelligent tools. Placement processes are faster. Firms have captured productivity and margin improvements.
Now what?
You can only remove so much cost. You can only automate so much administrative work. There are only so many points of margin to capture.
Cost optimization can create value, but it cannot be the long-term growth strategy.
To sustain organic growth and enterprise value, brokerages still have to win clients, retain them, deepen relationships, expand into attractive markets and build capabilities competitors struggle to replicate.
And that is where I think the AI conversation comes full circle.
When AI becomes broadly available, clients are unlikely to choose a broker because its copilot summarizes a submission slightly faster.
They will still ask the questions they have always asked:
Who understands my industry? Who has the strongest carrier relationships? Who can negotiate the best outcome? Who has the deepest expertise? Who can solve the difficult claim? Who do I trust when something goes wrong? Who consistently delivers?
Those were the foundations of competitive advantage before AI. I believe they will remain the foundations after AI.
The difference is that AI can reveal and amplify them.
A brokerage with stronger carrier relationships can use AI to surface market intelligence faster. A brokerage with deeper industry expertise can scale that knowledge across thousands of colleagues. Better producers become more productive. Better operations execute faster. Stronger service organizations become more proactive.
But AI also exposes the opposite. Weak data becomes harder to hide. Broken processes become obvious. Shallow institutional knowledge cannot suddenly be manufactured by an LLM. A technology layer cannot compensate indefinitely for weak execution or a culture that cannot change.
That is why the recent announcements from Marsh, Aon and WTW matter to me beyond the technology itself.
The interesting question is not simply what AI they are deploying. It is what they are connecting AI to: decades of placement history, claims experience, carrier intelligence, institutional knowledge, broker judgment and relationships.
Those assets took decades to build. AI can make them more accessible and more productive, but it did not create them.
Today, boards are asking: “What is our AI strategy?”
I suspect the more enduring question is: “What competitive advantages do we have that AI can make even more valuable?”
Because the short-term AI opportunity and the long-term strategy are not the same thing.
In the short term, AI can improve productivity, operating leverage and EBITDA margins.
In the long term, sustainable enterprise value still requires growth, differentiation and capabilities competitors cannot easily replicate.
Maybe we are misunderstanding AI when we talk about it as though it is itself the next great competitive advantage.
It may instead become one of the greatest productivity technologies our industry has seen.
The early movers should benefit. Margins may expand. Operations should get faster. Knowledge should become easier to scale.
But as those capabilities spread, competition returns to where it has always been:
Relationships. Trust. Expertise. Market access. Culture. Execution.
AI does not replace competitive advantage.
It reveals and amplifies it.
