Using AI in Your Agency: Risk Management Considerations 

Artificial Intelligence (AI) is rapidly changing the insurance industry. Agencies are experimenting with generative AI tools such as ChatGPT, Claude, Gemini, and Microsoft Copilot, while agency management systems and other insurance technology platforms are incorporating AI directly into their products.

While AI offers significant opportunities to improve efficiency, it also introduces new risks. While agencies do not need to avoid AI, they should adopt it thoughtfully, establish appropriate controls, and require meaningful human oversight.

How Agencies Are Using AI Today
Many agencies are already implementing AI to support operations such as:

• Drafting emails, marketing materials, newsletters, and website content
• Summarizing policy forms, endorsements, and carrier communications
• Comparing coverage forms and identifying potential policy differences
• Extracting information from applications, loss runs, and other documents
• Meeting transcription and note generation
• Customer service chatbots and virtual assistants

While these applications can increase efficiency, AI-generated output is not a substitute for professional judgment.

AI RISK MANAGEMENT TIPS

Education is Essential
Agency leadership and staff should understand the capabilities and limitations of AI technology.

Some examples of what AI systems can do include:

• Produce inaccurate information
• Misinterpret insurance terminology or coverage nuances
• Generate responses that appear authoritative, but are factually incorrect
• Omit important information or context

These inaccuracies can be particularly problematic when discussing insurance coverage, regulatory requirements, policy comparisons, or compliance obligations.

Employees should be trained to view AI output as a starting point, not the final answer. Information generated by AI should be reviewed and validated before being relied upon or shared with clients.

Develop an Agency AI Use Policy
Agency management should establish a written AI Use Policy that clearly defines acceptable and prohibited uses of AI technologies.

The following are some examples of what an AI policy should address:

• Approved tools and permitted uses
• Data security and privacy requirements
• Human review and employee responsibilities
• Client disclosure and regulatory considerations
• Record retention and incident reporting

Management should regularly review and update policies as AI technology, laws, and regulations continue to evolve.

Protect Data and Understand Privacy Requirements
Privacy and data security remain among the most significant concerns associated with AI adoption.

Depending on the platform and configuration, AI tools may collect, store, or utilize information entered by users. Agencies should establish clear guidelines regarding the use of AI with private, confidential, or proprietary information and ensure employees understand what information may and may not be entered into AI systems.

Before adopting an AI solution, agencies should understand:

• What information is collected
• How the vendor stores and protects data
• Whether customer data is used to train AI models
• How long information is retained
• What contractual privacy protections are available
• Whether customer data is segregated from other users

Agencies should also ensure their privacy policies and data governance practices contemplate the use of AI technologies and carefully review vendor agreements addressing data ownership, privacy obligations, security standards, breach-notification requirements, and data-retention provisions.

Additionally, agencies should evaluate whether carrier agreements, client contracts, or applicable laws place restrictions on how information may be shared or processed through AI systems.

Whenever possible, agencies should consider if enterprise-grade solutions provide greater administrative controls and contractual protections.

Use Caution with AI-Assisted Policy Analysis
One of the fastest-growing uses of AI in insurance is policy analysis and comparison.

Some AI applications can quickly summarize policies, identify differences between forms, compare renewals, and highlight potential coverage concerns. These capabilities can provide meaningful efficiencies when reviewing large volumes of policy information.

However, agencies should avoid treating AI-generated analyses as definitive.

Potential concerns include:

• Missed endorsements
• Incorrect coverage interpretations
• Failure to identify state-specific requirements
• Misidentification of policy changes
• Overlooking manuscript language or unique forms
• Incomplete comparisons between policies

AI can assist with the review process, but it should not replace professional analysis. Qualified personnel should review AI-generated recommendations before relying on them in coverage discussions, renewal reviews, or client communications.

The responsibility for advice and recommendations ultimately remains with the agency, not the AI vendor.

Be Transparent with Customers
Agencies utilizing AI-powered customer-facing tools should consider informing customers when they are interacting with an automated system.

Examples include:

• Chatbots
• Virtual assistants
• Automated service platforms
• AI-generated customer communications

Additionally, agencies should be aware that certain laws, regulations, or contractual obligations may require disclosure when AI is used in specific circumstances. As AI-related regulatory requirements continue to evolve, agencies should monitor applicable legal and compliance obligations and consult appropriate advisors when necessary.

If AI is collecting or processing customer information, agencies should be prepared to explain how that information is used and protected in accordance with their privacy practices.

Transparency helps ensure customers understand the role AI plays in the interaction and reduces the risk that AI-generated responses will be mistaken for individualized professional advice.

AI Is a Tool, Not a Replacement for Professional Judgment
The most successful agencies will likely be those that learn to leverage AI responsibly while maintaining strong professional oversight.

Clients rely on their insurance professionals to understand their exposures, explain coverage options, and provide guidance. AI can support those efforts, but it cannot replace the knowledge, experience, and professional responsibility of the agency and its staff.

Top

Risk Management Resource Highlight: Disclaimers

Disclaimers can be a vital tool in the agent’s toolbox to help you protect your agency against E&O claims. Our risk management website features a section of suggested disclaimers.

• Access the site by logging in to our online Customer Care Center and clicking the “Risk Management” link. Disclaimer examples are located under “Agency Tools.”
• CLICK HERE for instructions on how to sign up for and access Customer Care.