Managing general agents (MGAs) have been told there is an urgent need to establish formal AI governance frameworks before growing operational, cyber and regulatory risks become embedded within their organisations.

Speaking at the MGAA Annual Conference, cyber security and IT services provider Intersys said that while artificial intelligence is already transforming underwriting, claims handling and business operations, governance has failed to keep pace with adoption.

It added AI has enormous potential to help MGAs work more efficiently by automating data-intensive tasks across underwriting, submission handling and claims triage, allowing underwriters to focus on higher-value assessment and decision-making. However, Intersys’ research found that more than 80% of MGAs are already using AI, while almost half (48%) do not yet have a formal AI governance framework in place.

“The pace of adoption has been unprecedented,” It added.” While technologies such as Netflix and Facebook took years to reach mass adoption, generative AI tools such as ChatGPT have become mainstream within weeks, leaving many organisations struggling to introduce the governance, policies and controls needed to manage their use effectively.”

Without appropriate governance, Intersys warned in a new white paper that MGAs face increasing exposure to:

• AI hallucinations and inaccurate outputs
• Data poisoning and compromised decision-making
• Intellectual property loss and copyright breaches
• AI-enabled cyber-attacks and data leakage
• Regulatory and compliance failures
• Increased accountability risks for senior management operating within a highly regulated market.

Mark Kirby, Professional Services Director at Intersys, said: “Artificial intelligence is already changing the way MGAs operate and that’s something the market should embrace. Used well, it can improve productivity, streamline workflows and allow underwriters to spend more time applying their expertise where it adds the greatest value.

“The challenge is that adoption has happened much faster than governance. Too many organisations are already using AI without clear policies, oversight or accountability. In a highly regulated industry, it creates unnecessary operational, cyber and regulatory risk.

“The answer isn’t to slow down AI adoption. It’s to introduce the governance that allows organisations to use AI safely and responsibly. Firms that act now will minimise risk, protect their data and intellectual property, and be far better placed as regulatory expectations continue to evolve.”

The paper added unlike many sectors, MGAs operate under delegated authority, making underwriting and claims decisions on behalf of insurers while handling significant volumes of commercially sensitive and personal data. As a result, poor AI governance has implications not only for operational resilience but also for regulatory compliance, partner confidence and organisational reputation.

Rather than discouraging innovation, the white paper recommends organisations formally audit and document AI use, approve authorised AI platforms, provide employee training, establish clear governance frameworks and assign accountability for AI-related decisions. Together, these measures enable firms to capture AI’s benefits while reducing unnecessary risk.