Registered Insurance Brokers of Ontario issues guidance on responsible AI use for brokers

On May 29, 2025, the Registered Insurance Brokers of Ontario (RIBO) published the guidance “Responsible AI Use Among RIBO Licensees” to remind insurance brokers (collectively, Licensees) of their duties and responsibilities while incorporating the use of artificial intelligence (AI) in their business.

RIBO continues to examine how its mandate, guidance and existing regulatory requirements apply to Licensees’ AI use. RIBO acknowledges that there is an increasing interest in AI in the property and casualty insurance space and, while AI has many benefits, Licensees need to watch out for risks to consumer protection and professional standards.

AI use cannot replace a Licensee’s professional judgment or conduct. Licensees need to comply with the “Fair Treatment of Customers” and “Code of Conduct Handbook“ whether they use AI tools or not, including abiding by the following principles:

• Be competent;
• Act with integrity and in their clients’ best interests;
• Disclose any conflicts of interest;
• Protect privacy and consumer data; and
• Maintain client confidentiality.

RIBO encourages firms to develop an AI governance policy prior to using AI tools, and has set out the following four recommendations to assist with the development of such policies:

• Competency related to third-party tools and accountability for customer outcomes:
  • Licensees need to be trained to understand and identify when AI is being used and the risks of using AI within their business.
  • Licensees need to comply with the Code of Conduct even if they use third-party vendors’ AI tools to support or broker services that would otherwise be done by a person.
  • Firms should create governance structures that monitor risks and provide ongoing due diligence and education and training related to any AI-based tools and processes that are being used.
    
    
• Act in the client’s interest by proposing suitable policies:
  • Licensees need to oversee anything that has been generated or altered by an AI tool before presenting it to the client. RIBO uses the example of AI use in underwriting –firms should audit and monitor generated output to ensure that the model is working how it should be and is not affected by systematic biases.
    
    
• Human in the loop and transparent use:
  • Customers need to be aware that they are engaging with AI (e.g., chat bot) rather than a human – transparency is very important in this regard. Also, humans still need to be present to, among other things, respond to coverage questions or take corrective actions.
    
    
• Protecting privacy, consumer data and maintaining confidentiality:
  • Firms need to take all reasonable measures to ensure their customers’ personal information is protected while using AI. This includes developing policies
    1. To vet current and future vendors to ensure that any customer data processed through an AI tool does not leave control of the firm, and that it is not stored by the third-party vendor or used for training purposes; and
    2. That deal with authorized and unauthorized AI use and prevent unintentional privacy breaches from occurring.
  • RIBO iterates Licensees should not use an open AI system to process client information.

For more information on this topic, or in the event that you require assistance developing an AI governance policy, please reach out to the authors, Derek Levinsky or Claudia Lach, or a member of Dentons’ Corporate and Regulatory Insurance group for assistance.