Process is often described as the enemy of innovation. But in the rush to implement AI, the opposite risk is emerging: companies and leaders can become so focused on implementing solutions quickly that they lose sight of the responsibility bestowed upon them by their customers, agents, capacity partners, regulators and shareholders.
Good decisions require both
Start with the work that slows people down
There is a natural temptation to begin an AI initiative by asking what the technology can do and where you can place it within your business. I have found it more useful to start with understanding your
Where are employees waiting for information? Which tasks require the same manual steps every time? Where does a file sit for hours or days even though the work itself requires very little judgment?
As an example, at MSI we provide an insurance program for complex real estate investor portfolios. The process to quote an account once took as long as four or five days because of the manual work required to review the account, apply rating information and identify issues that needed attention. We used AI to develop a tool that supports faster ratings and generates alerts for further review. Most turnarounds have fallen to less than a day.
The value goes beyond a faster quote. Brokers receive answers sooner, underwriters can focus on the risks that require attention and fewer files are delayed by work that can be handled consistently through automation.
Every insurance organization has its own version of bottlenecks, manual work and friction. It may be policy intake, inspection review, claims documentation, renewal preparation, check processing or data validation. The best place to start is usually the task employees already describe as repetitive, slow or unnecessarily difficult.
Keep accountability with the professional
An organization is responsible for the quality of the output. The best way to maintain focus on this in an AI world is to keep accountability as close as possible to the professionals, decision-makers, process owners and functional leaders who own the deliverable.
A great example of this is our approach to underwriting, which involves far more than checking whether a risk fits within a set of guidelines. It requires context, experience and an understanding of how multiple factors interact. AI can prepare the information for that decision, but accountability should remain with the person making it.
We have applied that principle to high-net-worth home risk assessments. Reviewing the documents and comparing the information against underwriting guidelines previously took an underwriter approximately 20 to 30 minutes per file. AI can now complete much of the intake and comparison work within a few minutes. However, the underwriter still reviews the output, makes the final determination and is held accountable for the outcome.
That approach keeps human expertise exactly where it matters most. It also helps address a concern many insurance professionals understandably have, which is whether automation will gradually remove their role from the decision. We want to hand our best people better information and give them back the time to actually use their skills.
Measure what your best people do with the time back
Speed is an easy metric to communicate, but it does not tell the whole story.
Insurance leaders should also ask what employees are able to do once a process takes five minutes instead of 25 minutes. Are underwriters reviewing more complex accounts? Are they spending more time with distribution partners? Are teams identifying issues earlier? Has the quality or consistency of the work improved? As a leader, perhaps the most important question to ask your team is this: What new creativity shows up once your colleagues are freed from the manual and the mundane?
At MSI, we create new insurance products. While creative and interesting, the process of building a new insurance offering is also fairly manual. Insurance contracts need to be created, regulatory filings compiled, data files scrubbed.
With the assistance of AI, we are making the process more streamlined by automating the clunky, inefficient steps. As a result, our colleagues are spending more time building better, more comprehensive products and generating new ideas. Looking ahead, our goals are far more ambitious as we think about the types of solutions we can build for our customers and agents. That's what happens when we free up our colleagues' time and unlock their creativity.
Listen to the people closest to the workflow
Many of the strongest AI opportunities will come from colleagues who understand the day-to-day process in detail, not an outside consultant or an executive steering committee.
Colleagues know where information gets lost. They know which documents create delays. They know which steps require expertise and which ones simply require time. Leadership may set the direction, but the people doing the work are often best positioned to identify where AI can make an immediate difference.
To accelerate our AI initiatives, we created a group of AI champions at MSI, made up of people who already had an interest in the technology and direct exposure to the workflows we wanted to improve. Seniority mattered less than curiosity, practical knowledge and a willingness to experiment responsibly.
That combination helped turn general enthusiasm about AI into specific ideas the business could test. It also gave employees a role in shaping how the technology would affect their work.
Insurance organizations do not need every employee to become a technical expert. They do need a reliable way for good operational ideas to surface, get evaluated and move forward.
Build trust into the process from the beginning
AI adoption in insurance moves at the speed of trust. Employees doing the work, brokers placing the business, regulators overseeing the industry and customers whose coverage depends on it. They all have to be on board.
That requires real intention around AI-ready infrastructure, protection of proprietary data, adherence to regulatory requirements and regular evaluation of quality. A tool that performed well six months ago still needs oversight today.
Insurance professionals should be encouraged to question an output that does not look right. They should understand how the tool fits into the workflow and where their responsibility begins. Trust grows when people know that oversight is expected.
AI will continue to change insurance careers. The professionals who succeed will know how to direct these tools, evaluate their output and trust their own judgment when the answer isn't clear.
That is a valuable skill in any industry. In insurance, where every decision ultimately carries a responsibility to the people and businesses we serve, it's essential.
