Clean data and AI go hand in hand: Key takeaways from MGAA’s breakfast briefing with Azur Technology

The MGAA welcomed a fantastic turnout of members and industry leaders to its recent breakfast briefing with Azur Technology, held at the Lloyd’s Old Library. Discussions from the panel revolved around helping MGAs to cut through the Artificial Intelligence (AI) hype and unlock its real value, while maintaining human interaction.

Chaired by our CEO, Mike Keating, and sponsored by Azur Technology,  panellists included:

• Charlie Blackburn, CEO of Azur Technology
• Andy Hurrell, Founder and Head of Underwriting at Corin Underwriting
• Hasani Jess, CTO of Aventum Group
• Mark Mugge, CFO of Optio Group

As insurance firms increasingly integrate Artificial Intelligence (AI), it can often be a double-edged sword, equally bringing both opportunities and challenges.

On the one hand, AI has the ability to enhance operational efficiency, sharpen underwriting precision, improve the broker and customer experience, and create more structure in claims management, particularly for specialist and niche markets.

For example, it can significantly drive down costs, especially in cases where old legacy systems are still in use or outdated workflows. It also has the ability to implement something in a short time frame, lessening the time to market.

Staff were flagged as the most important asset, despite also being the largest expense. For firms that are trying to scale rapidly, AI can be a way of leveraging the business by maintaining data accuracy and transparency, and not having to increase headcount.

On the other hand, some concerns were raised about AI taking over lower-level jobs specifically, making entry-level careers somewhat harder to define and therefore creating a barrier for newly qualified candidates to enter the industry. But all panellists stressed that AI should be used to bring efficiency to the business by automating or modifying mundane, low-level tasks, with the purpose of freeing up staff to look at higher value work. One panellist highlighted that it depends on where firms are on their personal AI journey, pointing to the value of staff expertise for his MGA in particular.

But the success of AI is dependent on clean, structured data, yet many MGAs are hindered by outdated legacy systems, fragmented data in silos and manual processes. On top of this ongoing regulation and compliance around AI can prove tricky to keep on top of, especially when coupled with new cybersecurity risks and vulnerabilities where sensitive data is at stake.

Panellists agreed that compliance was particularly difficult to navigate across regions, with the UK having different regulations from say Europe, for example.

The emphasis on having clean and good quality data was a consistent theme throughout the debate, as well as making sure it was aligned with the business’s objectives. Therefore, data cannot be blindly used; accurate input of data is important as well as the right controls being in place where necessary, especially with the risk of that data travelling into the public domain – AI governance is key.

To counter this, actually measuring whether AI technology was allowing the business to grow exponentially was deemed tricky.

On top of this, data can often be fragmented and scattered, so efficient questions are key, and to this having a deep understanding of both your company’s core operational workflows is key. It is important for underwriters to have a good understanding of the data and to have it in one place, as scattered data across systems can impact the loss ratio.

Likewise, the pressure on the MGA sector to manage capacity in a soft market can make firms see AI adoption as a strategic imperative.

Another challenge around imagination and AI was flagged – in that until you have seen someone do something with AI, it may be unclear.

Overall, the sentiment was that AI is here to stay, but also that every firm is at a different stage in its AI journey.

The panel agreed that it was paramount to carefully balance the use of AI and human expertise, so the technology augments and does not replace human intervention entirely.

Panellists concluded that good data hygiene and the effective use of AI go hand-in-hand.