Building the Digital MGA of the Future
Keeping pace with a rapidly evolving insurance landscape requires more than incremental change. Technology and AI are now central to how insurers underwrite risk, optimise operations and deliver value to brokers and clients alike.
Since we last spoke with Alta Signa’s IT & Operations Manager Alex Cuadros, the rise of AI and automation technologies has accelerated dramatically, creating new opportunities – and new challenges – for specialty insurers and MGAs alike.
From reducing manual workloads to enabling more accurate pricing and faster decision-making, technology has become a key pillar of Alta Signa’s growth strategy.
In this latest spotlight interview, Alex shares his updated perspective on digital transformation in insurance.
What technological changes that you have recently witnessed are likely to have the biggest operational impact on underwriting?
We are now riding a wave of cognitive technology in the form of agentic AI platforms, which combine natural language with tools that connect to different context sources, so that an AI agent can orchestrate sub-agents and can understand when it needs to call on one or another of these sub-agents. So you can design that through a workflow and, within this workflow, teach any specific sub-agent to run a particular step within the process.
In the underwriting process, there could be a sub-agent which specialises in reading, interpreting and ingesting submission documents; another that opens the platform, creates a submission and fulfils it with the relevant data; another which queries everything it receives, flags any missing or inaccurate data, and then interacts with a human underwriter to highlight errors or request missing information; another that does the issuance; and a further agent which creates the monthly report and copies the information into the ledger – and virtually all of this without human intervention, but orchestrated by one agent communicating with the human underwriter.
This process completely changes the relationship between operating platforms and humans. The role of humans will be completely different in relation to technology, enabling them to focus their time on other high value activities (machine training being one of them).
In any case, these changes should be introduced gradually and within a framework tailored to each company. Leaders must carefully assess how any AI assistant fits the business, to identify the right approach, ensure a strong return and confirm that the necessary conditions are in place.
How should we balance technological innovation in underwriting, with retaining the human expertise specialty lines rely on and maintaining strong broker and carrier relationships?
AI would enable us to make our processes more automated and scale up our business in a much more relevant manner, but the output is not going to change. The machine isn’t going to establish or maintain the relationships with our brokers. We still need those personal relationships. We still need to listen to their needs, and ultimately, we still need to give them the service they expect from us. This is not going to change.
And while the agentic AI is kind of scary, you still can manage the natural language and the prompts that you are giving it, and you can also manage the tools. So, if you don't want a sub-agent connecting to your bank accounts, that's your decision. In the end, it’s a human who monitors and controls the right access to the right data and to any application. In my opinion, AI is not about replacement, but is about empowerment and augmentation. It can significantly increase human productivity.
What are the biggest technology challenges currently facing MGAs and specialty insurers operating across multiple European markets?
Two main areas of difficulty, which can take up a lot of time, are in the policy admin and the claims admin.
If you look at the MGA scenario, each time an underwriter assesses a risk and does the analysis and the pricing, he has to spend time looking at financials, corporate ESG scores, the company’s governance etc. With AI, this process can be simplified, so that more accurate and up-to-date information is acquired in a more timely manner, shortening the process.
The same is true with claims, when it comes to detecting fraud, or looking for information about one company or a subsidiary within that company. Sometimes the most up-to-date information is difficult to put together in a timely manner and you might have to invest a few hours in information gathering - now this can be reduced to just a few minutes.
Do you feel AI regulation in Europe is keeping pace with change?
Just like with other technologies, politicians and regulators have always been lagging behind AI. The changes in the market are going faster than we can create laws, particularly as we understand more of the potential problems once these technologies are implemented.
It is similar to social media – why are a few governments thinking about restricting access to social media to people younger than 16? Because, over time, we have seen real damage in the youngest people having unlimited access to networks, when they are not mature enough to discriminate between the kinds of information and messages that they are consuming.
But while legislators are always lagging behind, there is already an existing EU framework for data privacy, data security, risk management and specific directives about governance. However, in my view, at some point there will have to be deeper regulation of AI practices, data aggregation and AI’s extended usages.
Where do you think AI will have the greatest influence on underwriting and broker interactions over the next three to five years?
That’s very difficult to answer, because change is happening so rapidly. Perhaps we could create a virtual MGA where the platform picks up all your policy submissions and where sub-agents perform all the policy and claims servicing. But if you had asked me that question three or four years ago, I wouldn't have imagined we could automatically handle the entire process through sub-agents!
If I look ahead six months, we are on the path to creating agentic AI platforms capable of taking out much of the burden of underwriting and claims processes, under certain conditions and where applicable, so human resources can be focused on higher-value tasks: more accurate and timely data gathering, better risk analysis, more accurate and segmented pricing.
AI is about technology and technology serves the business of the company. That means that it must be used to leverage profitability through increasing a company’s productivity or by helping to find new business models that are data intensive. It is important not to forget that.
_______________
Thank you, Alex, for taking the time to speak with us in such a fast-paced world. Your insights offer a fascinating glimpse into how AI and automation are reshaping insurance, while reinforcing the value of human relationships and expertise.