AI agents: a leap forward in the insurance space

Readingtime: 5 min

“We are the first to shift our entire platform onto artificial intelligence (AI) agents,” said Arndt Gossmann, chief executive officer of DGTAL, emphasising their groundbreaking potential in the insurance sector.****

“AI agents are a major leap in generative AI (Gen-AI), moving beyond generating text or presentations to performing more complicated operational tasks,” he said.

Gossmann’s enthusiasm for this technology is contagious, but what exactly are AI agents?

Picture a team of mini-AI helpers, each assigned specialised tasks, working together to support human decision-making.

“AI agents are essentially little helpers,” Gossmann explained. “Instead of having one big software running entire processes, the process is delegated to a team of AI agents, each of them equipped with software, access to sources and specific abilities.

“AI agents coordinate and communicate among each other to plan and execute all steps to complete a process. Obviously it happens very fast, but step by step and it is fully transparent. The agentic approach also allows interaction with human agents or the allocation of AI agents as a perfect support team for the human expert.”

These agents then collaborate to solve complex issues.

“They can tell each other: ‘I have an email I can’t handle or a question I am not sure about ’ and either reassign it to a different AI agent or get help from the human expert in order to respond best.”

AI agents are a completely new concept to use Gen-AI. For insurance it is a very big step forward, and Gossmann puts their transformational potential on a par with the industrial revolution.

“For information processing industries, especially insurance, AI agents are a game-changer.

“They can plan, use tools, execute, and delegate tasks among themselves. They aren’t just following a script—they’re able to adjust and react based on what they encounter.”

This kind of autonomy is a huge boost in operational efficiency.

“Insurance is complex,” Gossmann admitted. “Although AI agents can’t take on the full processing cycle themselves just yet, they can help users find what they are looking for and handle the tedious parts, freeing up human experts to focus on higher-level decision making.”

One of the key areas for AI agents is in claims processing, particularly in response to natural catastrophes.

“After a flood, the influx of claims is overwhelming,” Gossmann explained.

“AI agents can step in to assess and process these claims much more quickly than humans. After large events it takes weeks until the insurer has a clear picture. Following Hurricane Milton the estimates still vary by billions week by week. AI can provide a daily measure instead of a weekly estimate. This is highly relevant for capital providers such as insurance-linked securities.”

In this context, AI agents also assist with faster claims settlement, ensuring that customers get the support they need without unnecessary delays.

“The damage might have already occurred, but the more quickly the insurer can act, the better the client journey will be,” Gossmann asserted.

Useful across many disciplines

Gossmann identified three other areas where AI agents can have a transformative impact: underwriting, customer experience and decision-making.

“Each AI agent improves with time. It’s not just that the whole system becomes more efficient—the individual agents themselves learn and adapt,” Gossmann said.

This capacity for continuous improvement makes AI agents invaluable in tasks that require ongoing refinement.

“The first question when wanting to utilise AI should always be ‘what are the challenges?’. AI should be applied at the pain points,” Gossmann said.

He pointed to flooding in Europe as a key example. “Flood claims come in waves, and they’re overwhelming for insurers.

“AI agents can take over the first level of claims notifications, flagging important details and freeing up human staff to focus on problem-solving and customer interaction.

“We don’t want experts to waste time reading through endless documents. We want them to focus on the vital tasks, such as assessing information, making the decisions or developing new insurance products for fast emerging risks.”

“They combine the strengths of both AI and human expertise.”

By putting a team of AI agents around each expert, companies can operate far more efficiently.

“You can have AI agents acting as research assistants,” Gossmann noted. “For example, an expert might want daily updates on hurricane developments. Instead of trawling through data, they just instruct the AI agents to do it.”

Despite their capabilities, Gossmann is clear that AI agents aren’t meant to replace human capabilities but augment them

“AI agents are co-pilots: when they encounter something they can’t solve, they pass the task on to a human. That’s why they are such a powerful tool—they combine the strengths of both AI and human expertise.”

Looking ahead, Gossmann is confident that AI agents will become a cornerstone of insurance operations. “

This is the future,” he said. “It’s not a black box any more: you can trace how the agents are working, see which tasks they’ve taken on and monitor their progress.”

The promise of speed, quality, and collaboration—along with the ability to learn and adapt—positions AI agents as an essential tool for the future of the industry.

As Gossmann put it, AI agents represent a “pioneering way” to use AI in insurance.

Arndt Gossmann is chief executive officer of DGTAL. He can be contacted at: a.gossmann@dgtal.io

For more news from Baden-Baden Today, click here.

Fitting to the topic: