Embracing AI: OPC Guidance on Generative AI

Jon Davies
30-May-2023 16:12:08
Artificial intelligence (AI) is a popular topic this year and is steadily becoming a mainstay of business operations; we're certainly excited about it as a team.
 
New Zealand's Office of the Privacy Commissioner (OPC) has recently released essential guidance on the use of generative AI. See here for the recommended guidelines. Full disclosure: At InsuredHQ, we are assessing the deployment of generative AI for our customer support services, and we greatly appreciate the practicality and relevance of the OPC's guidance in this context.
 
 
Generative AI tools, such as OpenAI’s ChatGPT or Google's Bard, have the potential to significantly enhance customer service experiences by providing personalised, real-time responses. However, their use does come with potential privacy risks and the OPC's guidance offers insightful recommendations we intend to incorporate as we develop our strategic approach to generative AI.
 
Here are some key points from the OPC's guidelines and how InsuredHQ is planning to incorporate them:
 
  1.  Senior Leadership Approval: We will ensure that our senior leadership team is fully informed of the risks and rewards of deploying generative AI and has given explicit approval.

  2.  Necessity and Proportionality: We are committed to assessing whether the use of generative AI in our operations is both necessary and proportionate to the intended benefits for our clients.

  3.  Privacy Impact Assessment: In line with the OPC's guidance, we will conduct a comprehensive assessment before implementing any generative AI tool to identify and mitigate potential privacy risks.

  4.  Transparency: We will keep our customers informed about using generative AI tools in our customer support system. We will explain how, when, and why these tools will be used and how potential privacy risks will be addressed.

  5.  Procedures for Accuracy and Access: We will develop procedures to ensure the accuracy of the personal information collected and provide a transparent process for our customers to access and correct their information.

  6.  Human Review: To mitigate the risk of inaccurate or biased information, we will ensure a human review of the outputs generated by the AI tool before any action is taken.

  7.  Data Security: We will ensure that the AI tool does not retain or disclose personal or confidential information, aligning with the OPC's recommendation. 
The recommended guidelines issued by the OPC provide a comprehensive and thoughtful approach to integrating generative AI into organizational workflows. At InsuredHQ, we remain committed to delivering top-notch service to our customers while protecting their privacy. Therefore, the OPC's guidelines serve as an invaluable resource for our journey towards implementing generative AI in our customer support services.
 
As always, we strive to stay ahead of the curve and continually look for ways to enhance our customers' experience while maintaining their trust and privacy. Our careful, measured approach to its implementation, prioritizing privacy and customer interests, will set a positive example for the wider industry.

Subscribe by Email

No Comments Yet

Let us know what you think