ChatGPT - How To Use Within Enterprise?

ChatGPT is the talk of the town across all corners of the world. In short, it is an artificial intelligence chatbot that leverages natural language processing to generate human alike conversational dialogues. It was launched on 30th Nov-2022 by OpenAI.Users can have access to the GPT-3.5 model freely. Paid subscriptions are required to access the more advanced GPT-4 model.

After it was launched, people are debating for its possible use at the enterprise level. We need to be prepared before starting to use in industries to reduce the risk from the operation, information security, and privacy standpoints. We need to think about the following as part of our preparation strategy before we start using them in industries -

1. Enterprise-level AI( Artificial Intelligence) policy needs to be formulated which shall divide ChatGPT use cases into three categories-

a > Use cases that are prohibited (e.g., using ChatGPT to check for mistakes in confidential company or client documents, or sensitive company code)

b > Use cases that are permitted with authorization from some designated authority (e.g., generating code, so long as it is carefully reviewed by an expert/panel of experts before being implemented)

c >Use cases that are permitted without any prior authorization (e.g., creating purely administrative internal information such as generating ideas for icebreakers for new hires)

2. Deployment of AI risk governance framework that includes the following touchpoints at a minimum-

a> Creating a set of criteria to understand whether the ChatGPT use case is low, medium, or high risks.

b> Maintaining a proper inventory tagged with a particular team which will keep track of the use cases and evaluates them as low, medium, or high risk based upon the established criteria.

c> Internal labeling is required so that the output for ChatGPT is marked with an identifiable marker so that the reviewer pays special attention to those.

d> Notifying clients/ external world if the contents are generated by ChatGPT and shared with them.

e> Providing periodic training to employees on acceptable and prohibited use of ChatGPT( substantiated by enterprise-level AI policy)

To summarise, ChatGPT is the beginning of a revolution that will surely transform our lives in the future. However, we need to put guardrails around so that we start using it effectively, judiciously, and ethically bound by regulations, policies, and frameworks