Review: Mastering Crisis Communication with ChatGPT

About the author

Richard Bailey Hon FCIPR is editor of PR Academy's PR Place Insights. He teaches and assesses undergraduate, postgraduate and professional students.

Mastering Crisis Communication with ChatGPT:
A Practical Guide to Writing Perfect Prompts and Communicating Effectively in Any Situation
Philippe Borremans
2023, 67 pages, Kindle ebook

With the exception of a few specialist consultants who could be considered full-time crisis communicators, most people working in public relations and comms encounter full-blown crises infrequently. The shock of experiencing one is that you’re suddenly in the spotlight and facing multiple simultaneous demands. There’s so much to do and little time to do it in. 

It’s in this context that this small ebook could be career-enhancing. It shows how to use ChatGPT as an efficient and unflustered assistant. Even better, it shows you how to plan ahead.

This sensible, practical, clear and timely book from a respected international communicator and trainer balances the risk and opportunity equation we face when considering how to use artificial intelligence. While a large language model such as ChatGPT can be used to create ‘draft responses, FAQs, social media posts and more during a crisis’, Borremans notes that ‘[ChatGPT] may not always be able to understand the nuances of a crisis scenario and the context in which it occurs.’ Humans still needed!

Besides, ‘it raises ethical concerns. There is a risk that it could be used to generate responses that are misleading or harmful, or to spread misinformation in a crisis scenario.’

Some might be put off at this point, but since there’s an evident danger you could face a crisis situation because of the use by others of AI to spread false narratives about your client or organisation, then you would be wise to be forewarned. 

And do get over the worry that ChatGPT is about to put you out of work. There aren’t any juniors who are battle-hardened crisis communication experts, and it’s the junior tasks involving research and copywriting that are most immediately threatened by generative AI. Yet senior practitioners, those more likely to be faced with a crisis situation, are also perhaps those less likely to spend time playing with new tools. So they should read on.

ChatGPT is a powerful tool, but it should not replace human judgement. The final decision on how to respond to a crisis should always be made by a human. Philippe Borremans

Borremans advises us that if we want better answers from ChatGPT we need to ask smarter questions. He offers one practical tip for achieving this: the ‘act as’ hack. You could instruct ChatGPT to ‘act as an expert’ in order to generate an authoritative statement or to ‘act as a journalist’ to create some news-style articles or tough questions.

Early on, he asserts the continuing need for a crisis communication plan. ChatGPT does not and should not remove this requirement. Yet it can be used for testing the plan by generating some fictional crisis scenarios. And here it gets smarter because the author tells us to use ChatGPT to create prompts for ChatGPT – on the grounds that it knows the language to use to get best results. He shows us how this works for a simulated cybersecurity crisis, and he praises ChatGPT for considering internal audiences and for proposing a post-incident review.

Having created the fictitious cybercrime scenario, Borremans shows how to prompt ChatGPT to draft a reactive statement. Then how to rewrite it to demonstrate transparency and build trust in the organisation, a step he calls ‘human optimisation’.

You may not choose to use the text, but it only took a few seconds to generate so can be a time saver and useful support for the communicator facing a crisis.

So you now have a reactive statement, or a statement optimised for internal audiences, composed in intelligible English. Borremans then uses ChatGPT to request a French version – as it’s another of his languages so he can check the results. He finds it ‘as good as perfect’ but does warn that you should always check a translation with a native speaker who also understands the ‘tone’ of the organisation.

Borremans also shows how ChatGPT can be used for ‘Red Teaming’ – for anticipating future risks and for developing plans against them. What questions are business journalists likely to ask? Check with ChatGPT, and prompt it to ask aggressive questions.

Need help preparing your crisis plan? He shows how to produce an outline with the help of ChatGPT.

He ends by asking whether AI could replace our jobs and concludes with this unsourced quotation that’s being widely cited:

You might not be replaced by AI, but you could definitely be replaced by another professional with AI skills. Anonymous

Don’t wait to take notice until you’re facing a crisis. By then it will be too late.

A note on format and pricing. I paid £8.79 through Amazon to download this Kindle ebook (you can view it online if you don’t have the Kindle app or reader). You can read it in an hour or two. So looked at one way, it’s quite an expensive book. Looked at another way, it’s a very cheap training course from an expert communicator written for fellow senior practitioners who may not have found the time or motivation to explore the potential of ChatGPT.

Please also note that this ebook was prepared using the free version of ChatGPT. A more capable subscription version, ChatGPT 4, is now available.