Newfront has been iterating on a proprietary insurance workflow engine and structured data model for more than six years. From the beginning, we have found ways to apply machine learning and predictive analytics to solve our clients’ needs. With the recent explosion in generative AI, Newfront became the first AI-powered brokerage and introduced an ethical framework, called AI Principles, to ensure responsible build and implementation. Newfront’s Chief Technology Officer Gordon Wintrob and Head of Data Patrick Miller announced the principles internally earlier this summer, a first in the world of insurance brokerages and AI. With these guardrails in place, the Newfront engineering team partnered with our insurance experts to launch solutions such as the AI Benefits Assistant, a contract review tool, an instant way to distill thousands of audit reports and 5500s, and more. 

Here are the full Newfront AI Principles that guide our approach:

At Newfront we deeply believe that the recent advances in generative AI present a massive opportunity for the insurance world and for our company. AI has the potential to significantly unlock our insurance professionals’ time for more meaningful, client-facing work. When used hand-in-hand with the top brokerage talent at Newfront, we’ll be able to deliver deeper insights to clients about their business risk and total rewards.The advancements in AI also present clear risks if not deployed responsibly. To that end, we are declaring a set of principles that we will follow for all of our AI-backed products and services:

1. Favor human-in-the-loop AI systems and leverage our insurance professionals’ expertise. We leverage our deep insurance expertise both in building the AI systems and in delivering AI-driven insights to our clients. When we deliver AI products directly to our clients, they have been thoroughly vetted by our insurance professionals.

2. Ensure security and privacy of our clients’ data. We develop data products and AI solutions with an eye towards making sure our clients’ data is secure and not shared in unintended ways.

3. Be thoroughly tested before being deployed. We comprehensively measure performance and ensure safety of our AI systems and applications before we deploy them to production.

4. Avoid reinforcing unfair biases. Insurance should be accessible to all, and we strive to build data products that can work for all clients.

5. Incorporate transparency and explainability. AI doesn’t need to be a black box. We build AI solutions that can be explained and we disclose when it is being used.

All of our externally-facing AI products get formal AI Principle reviews to ensure that what we develop is accountable to these principles. At Newfront, we are uniquely positioned to deliver human-in-the-loop AI solutions. Not only do we think this is the responsible way to innovate, but it also fits perfectly with our ethos and mission of supercharging insurance brokers.