We’ve been studying artificial intelligence (AI) and its use cases for a multitude of scenarios. “Generative AI” is quite popular these days in the form of large language models (LLMs), including ChatGPT and Claude.
For us strat comms/PR/IR folks, LLMs can be helpful with tasks like summarizing notes and testing writing for sentiment.
But there are also inherent dangers.
For example, let’s say you’re using ChatGPT to clean up notes you took during a meeting where material non-public information (MNPI) was disclosed, or you’re involved in a lawsuit, and you prompt an LLM with information that somehow informs opposing counsel of your strategy.
As soon as you submit a prompt to one of these platforms, you are effectively disclosing information.
We talked about these issues, as well as putting a formal AI policy in place at your company, during a recent NIRI webinar on how to use AI safely and effectively.