In this next "bake-off" between ChatGPT and Hailey, 6clicks' AI assistant, we aim to highlight the differences between a general public AI model and a tailored AI model in terms of answering queries that require enhanced context and authorization awareness.
As we have initially established, Hailey Assist is the world's first AI assistant for GRC, specifically designed to connect organizations with relevant information and real-time insights into their GRC programs. Hailey takes the form of a chat interface that is accessible to users of the 6clicks platform. Some of Hailey's capabilities include answering questions about your GRC program, creating risks and other tasks specific to the 6clicks platform, and helping users easily access items and modules across the platform.
For today's video, we will see Hailey go against ChatGPT in providing context and authorization-aware responses:
Upon testing both AI models, my first question was "Should we encrypt data to 128-bit or 256-bit?" As expected, ChatGPT provided a general response. It didn't consider any specific context about what data is being encrypted or why and simply explained the basics of encryption standards without any organizational context.
On the other hand, for Hailey, I switched to a user account with reduced access to only their organization's controls within the Controls module of the 6clicks platform. When asked the same question, Hailey's response was specific and context-aware: "Logistics data encryption is implemented using AES 256," and it provided a link to the relevant data encryption control. Hailey was able to reference the organization's policies in its response and provide accurate information based on the organization's specific context, demonstrating a significant advantage over ChatGPT.
To further illustrate Hailey's capabilities, I asked another question: "Do we have any overdue assessments?"
Given my restricted user access, which only includes controls and not assessments, Hailey couldn't respond with any data about assessments. This limitation is a testament to Hailey's robust permission and authorization model, ensuring that users only access data relevant to their roles.
In essence, while ChatGPT offers broad and generalized answers, using a tailored AI model like Hailey for organization-specific queries provides you with context-rich, precise responses that align with your organization's unique requirements and permissions.
Leverage our AI assistant, Hailey to instantly retrieve critical information about your GRC program, educate your employees on security policies, and harness insights such as your risk posture and compliance status to improve decision-making. For a complete experience of Hailey's capabilities, book a demo with us below.