The Future of the Agency in a World of AI

By Andrea Wells

Technology is changing the world at a faster pace than ever before with the help of artificial intelligence (AI). And the sky is the limit on what AI can and will do for the future of the independent agency system.

“The sky is the limit,” said Ryan Hanley, president of Linqura AI and the founder and president of Finding Peak. Every technology that’s come along promising greater efficiency and positive change for independent agencies has served its purpose. Early adopters of those technologies were rewarded, but eventually even late adopters, laggards, found their way to the same reward, he noted. “Everyone rapidly caught up and then there was very little lost space between the two,” Hanley said.

But with AI adoption, the agencies that are building a culture around these innovative technologies will leave the rest behind, he stated. “Those companies are going to ramp up so fast that when the laggards start to adopt this technology, they’re going to be so far behind that they will not be able to catch up.”

The power of AI is limitless, Hanley believes. “There’s no ceiling,” he said. “We don’t know what the ceiling is today. So, the more expertise, time, and capital that you can apply to it, the more benefit you will receive from it,” he said. The growth and ability to do more with increased efficiency and scale is exponential, he added.

“For example, I could have Sonnet AI answer all my phone calls and direct those phone calls to the appropriate person, and where necessary collect information after hours or on weekends or for specific programs in order to quote business without ever talking to an actual human being–and do that with 95% accuracy and only a 3% complaint rate,” he said. “That alone, if you’re implementing the AI properly, changes the entire nature of your agency.”

Hanley said there’s a significant gap in knowledge on how best to use AI in agencies. Someone who is just beginning to use a free ChatGPT account to write a few LinkedIn posts is already behind in the AI revolution, he said. “They are using AI, yes, but they are so far behind, it’s not even funny,” he said. “And there is an entirely different group of agencies that hasn’t even touched AI’s capabilities yet,” he added.

For many agencies, the age of AI is moving so fast it is difficult to keep up and understand how and where to use the most effective tools. In recent years, insurance agencies have been overwhelmed with the amount of technology coming to market, particularly in the AI space, said Kasey Connors, executive director of the Big “I’s” Agents Council for Technology (ACT).

“Where do they start, how to adopt it, ease of use, risks, and responsible change management” are issues agents ask about, Connors said. “Agencies are really looking at AI right now and asking how to get started, and then how to work with vendors or vet them, and then how to integrate those AI tools into their businesses,” she said. “As with any technology, there’s a people change management piece and then how to stay current with that change,” she said.

Connors also agrees with Hanley that the knowledge gap and use of emerging AI tools in the independent agency world is significant. “In talking to many agents, some are all in–they are using it for direct bill reconciliation, they’re using it for policy comparisons, they’re integrating it into their marketing, they are leveraging it for data, or voice assistance,” she said.

‘There’s no ceiling.’

The ways in which agencies are using AI tools vary significantly, she added. The agencies that are seeing the most return on their investment in AI are using the tools to solve real problems, Connors said. “It’s one thing to leverage ChatGPT to help with an email versus integrating a tool that is solving for operational efficiencies. I think it is critical for agencies to understand that difference,” she said. “It’s not just adopting AI for adopting AI’s sake; it’s really about how can you think about this tool for your business.”

GPT-ing in the Agency

ChatGPT has been a game-changer for many businesses–from streamlining workflows to helping with customer service. Many small businesses today even use it to help onboard and train new employees. But one area where agents may find it useful is helping to sort vast amounts of information–even policy information–in seconds. That’s why Relation Insurance Services embarked on structuring its own RelationGPT this year.

Kevin Rabinowitz, chief operating officer at Relation Insurance Services, told Insurance Journal that RelationGPT was built on OpenAI’s ChatGPT 5 model, tapping into the significant investment that OpenAI made in that model. But while OpenAI is public-facing, RelationGPT was designed to keep information withing the ecosystem of Relation Insurance Services.

Security of client information is a top concern, Rabinowitz said. “So, the way we have this set up with OpenAI is it’s not just like a straight license of their product where they own the data that goes into it,” he said. “We’ve actually built our own; we call it an orchestration layer that sits in between our information and ChatGPT … It’s kind of a pass through, but it’s HIPAA compliant,” he said. “It’s enterprise safe, and our information never leaves the walls of Relation, so we feel really confident in the security of the system.”

Rabinowitz said employees are able to use the AI tool in a number of ways, similar to how an independent user might utilize ChatGPT but tailored for insurance needs.

“We’re able to use it and deploy it internally in an enterprise-safe manner that helps us be more efficient and better serve our clients,” he said. “So, it looks like ChatGPT, and it’s fully equipped with capabilities around document comparison, harnessing information that’s available on a variety of sites to improve our team’s knowledge about different insurance topics and coverage related topics,” he explained.

Document comparison is one of the primary ways that brokers use the technology, Rabinowitz said. “Sometimes that’s a comparison of two different policy forms. Sometimes it’s a comparison of contracts. Sometimes it’s a comparison of an insurance quote to a policy. It really runs the gamut,” he said.

So far RelationGPT has proven to be an amazing time saver for employees who used to have to compare every word, he said. “Imagine comparing a 40-page contract in seconds,” Rabinowitz said. Efficiency, speed, and the quality of the agency’s internal operational process has been the focus of the development of RelationGPT, he noted. It is helping to connect the steps in many internal processes that are needed to successfully serve clients, he added. “There’s just so much opportunity in the brokerage segment for us to improve the quality and efficiency of our own internal processes using technology like this,” he said.

For Relation Insurance Services, the challenge was not creating the tool as much as creating the change for people. Incorporating technologies like AI and driving adoption among people who aren’t used to these kinds of tools can be a hurdle to overcome, Rabinowitz said. “That’s been a lot of the fun that we’ve had this year–sorting through those adoption and digital readiness considerations,” he said.

But there is no other option than to move forward with technology tools that can help to improve efficiency within the organization and for its clients, Rabinowitz said.

Next Generation AI

The next generation of agency AI tools will be focused on doing more tasks without the need for a human. This new generation of tech, known as agentic AI, will be an “agent” that mimics the work of a human. Agentic AI tools will have the capability to understand and respond to customer inquiries without human intervention.

‘There’s just so much opportunity in the brokerage segment for us to improve the quality and efficiency of our own internal processes using technology like this.’

Imagine the possibilities. Imagine the risk. How will AI handle the exposures that might come from an agency error or omission in coverage? But some innovators in the AI development world say AI agents will only help to reduce the risk of error in agencies.

San Francisco-based SuperAgent AI founded this year says its artificial intelligence agents can handle insurance advisory, sales, and customer service at all hours of the day–more efficiently than traditional agents and with fewer mistakes.

“What Uber did for transportation, SuperAgent will do for insurance,” said Milan Veskovic, founder and CEO of SuperAgent AI.

“Our fully autonomous AI agents will eliminate human error, offer superior client interactions 24/7, and fundamentally alter industry expectations,” according to the company. “As the first real-time AI co-pilot built specifically for insurance sales teams, we empower agents to perform at their best, on every call. Our platform helps agencies ramp new hires faster, boost close rates, and bring consistency to every conversation,” the company says.

Veskovic admits that there are a few barriers to overcome to get to a fully autonomous agent, including state licensing regulations for AI agents. “We are very early on in this journey,” he said. “But we want our insurance agents to be licensed in all 50 states.” Veskovic noted that the firm is currently in discussions with two major regulators about this issue.

For now, SuperAgent is working with agents as a co-pilot solution, explained Vlada Lotkina, chief operating officer at SuperAgent AI. Lotkina said the AI system’s methodology is broad and focuses on learning from real conversations about humans.

“It’s deeply rooted in behavioral and conversational dynamics, which allows it to really be the expert at insurance sales first, and then the product knowledge is the layer on top of it,” she said. That allows the system to be flexible and agile and effectively grow with the needs of the agency, she explained.

Lotkina said the co-pilot solution requires limited integration requirements, if any at all, with other agency technologies such as agency management systems. “That allows us to really learn with the organization, grow with the skills, while helping the existing agent become more productive, leveraging our technology,” she said.

At the same time, SuperAgent AI is also feeding its own algorithms or proprietary methodology that will eventually allow those agencies to launch fully autonomous AI agents in the short future, she said. These AI agents will serve a role in helping “the people that they have on their teams.”

For now Veskovic said that the sales platform is helping human producers by giving them the support they need to succeed. “We are increasing cross and conversion rates for existing agents, increasing the number of policies per household, helping them more efficiently go through the sales conversation, and with analytics, reshaping how insurance sales managers are managing teams,” he said.

“You will not only know approximately what your team is doing and what it is that they delivered at the end of the month, but you will know exactly the quantity and the quality of their activities,” he said. That process also creates a self-learning loop for the SuperAgent AI tool, he added.

When it comes to a future of fully autonomous AI agents that provide true quoting and binding of new business, there are hurdles to overcome, including integration with outside tech systems, regulation, and potential E&O exposures.

“We are very early on,”

Veskovic said. “Yes, of course we have this in mind, but the thinking is simple,” he said. “A human makes mistakes,” he noted. “Our solution makes less mistakes for sure.” Even so, the final review of a “report” on activities will be a human. “So, as a human, I can oversee what my tool did, and then maybe call the customer and say, ‘Hey, I have to make a small adjustment.'” But because of the AI’s self-learning loop, Veskovic said he doesn’t consider these experiences to be a “great threat” when it comes to E&O exposures. “The insurance agent is the one responsible for delivering the solution to the client,” he said. The agent holds the license to sell the insurance product, he added. At least for now.

Veskovic hopes to change regulatory requirements that will allow AI agents to be the selling agent some day. “So, while we are changing the legislation, we will take on the responsibility to make sure that we execute everything properly and we will be compliant in all the 50 states,” he said.

For now, “we are collecting conversational and behavioral data, completely anonymized, in order to improve the loop and improve the performance of our solution to serve the customer better and to serve the agency better,” Veskovic said. “That allows us to match what the needs are–not just price, location, the base information, but really understand the deeper context that today, outside of SuperAgent, just does not exist in insurance sales.”

Human vs. AI

So, the big question is will human agents continue to be needed in the future of the agency system?

Rabinowitz thinks yes. “What we see in terms of the consumer sentiment, what clients like when they’re purchasing insurance is to have a trusted source that they can rely upon and ask questions,” he said. Even in areas such as personal lines or small commercial. “There’s always going to be a need for that when a client is purchasing insurance, to have someone that they can rely on to ensure they’re getting the right information and making the right decision,” he added.

‘The independent agency channel is relationship focused, and I don’t see people fully 100% going away.’

“What we need to really focus on is using technology to augment the quality and strength of those human-to-human relationships that we build with our clients and not lose sight of that,” Rabinowitz added. He said it’s important to think about incorporating technology into that process of building and maintaining a relationship. That’ll be the challenge that agencies will face in the future, he said, striking the right balance between automation and people. “But not losing the special quality of having a relationship with a client that keeps them coming back and keeps the trust there.”

ACT’s Connor thinks the agencies that are adopting AI successfully right now will see a shift in what’s needed from their human workers. “It’s freeing up time for others so that they can still focus on what they do best, and that is the relationship business of talking to clients and helping them from that advisory capacity,” she said. “The independent agency channel is relationship focused, and I don’t see people fully 100% going away.”

SuperAgent’s Veskovic sees a big transformation in agency roles as a result of artificial intelligence in the future. “I’d say the job as they know it now will be eliminated,” he said. “It’s just a matter of time.”

Hanley agrees with that prediction and more. He thinks the era of maintaining a personal relationship with clients as it is today will eventually come to an end. People will grow accustomed to using AI voice agents and chat functions in insurance as they are already used to using them in other areas of their lives, he said. “I think that the narrative (of personal relationships), while it’s been true for a while, is coming to an end.”

Hanley also knows that is not a popular thought among the industry. “There are a lot of agents who will say things like, ‘Well, we want to maintain our personal relationship.’ That to me is, and I’ve always said this, and I’ve caught a lot of heat for it, is very selfish thinking,” he said. “I think a lot of independent agents apply the way they want to do business with their clients versus understanding the way their clients want to do business and then adapting to it,” he said.

But in reality, Hanley added, no one really knows how much the industry will change as a result of AI in the future. “We have no idea. No one can predict it. I can’t predict. All I know is everything is going to be different,” he said.

Who knows, maybe there will be a counter movement back to humans, Hanley said.

“Do we get so fed up with AI voice that there’s this counter move to want to talk to a human, and maybe (consumers) pay more for that? Are there going to be VIP white glove agencies that say, ‘Hey, you can pay an additional fee to work with a human versus work with an AI agent?”

It’s truly the Wild West of agency AI, according to Hanley. At least for now.

Topics Trends InsurTech Data Driven Artificial Intelligence