Generative AI and large language models (LLMs) are transformative tools reshaping nearly every facet of the working world. This includes a significant democratization of communications. Using AI, a number of businesses have taken to developing press releases, creating blog posts on topics relevant to their audiences, and producing communications plans to help them navigate myriad situations, including moments when a crisis strikes.
One would think this is a moment of existential dread for communications professionals if these tools can take over the work we do on a day-to-day basis. Why should companies hire communications roles or engage outside PR agencies if ChatGPT can do this work faster and cheaper? This is a valid question. But, before we assume AI will fully replace communicators (despite Microsoft listing Public Relations Specialists as number 23 of the top 40 most affected occupations by generative AI), let’s take a moment to look at what AI can do well and where it falls short.
Where AI Excels in a Crisis
As an experiment, I prompted ChatGPT as if I were the CEO of a mid-sized company manufacturing engine components and had recently discovered my CFO was embezzling from the company for several years. The fraud involved stealing from the company’s general fund, inflating contractor fees, and skimming from employee bonuses. I asked it to create a plan to help me communicate about the situation to my employees and customers.
I won’t lie, what ChatGPT was able to produce based on that minimal prompt was solid. It provided specific priorities with recommended timing, a high-level internal communications strategy with key points to cover, follow up actions, and more.
All in all, what AI put together is fairly consistent with what my colleagues and I would advise any client that reached out to us with a similar issue.
The Critical Gap: Empathy
However, I did notice one important point ChatGPT missed which, in my opinion, is the most crucial element of all crisis communications: empathy.
Not once in the plan I received from ChatGPT did it mention or focus on empathy. Even though two components of my prompt indicated specific groups were impacted (contractors with inflated fees and employees with skimmed bonuses), there was no guidance on recognizing their frustration, anger, or sense of betrayal.
That is a significant miss, and likely why a study from Corporate Communications: An International Journal (as reported in PsyPost) shows “human-written crisis messages are perceived as more credible and reputationally beneficial than those authored by AI systems.” Audiences don’t just evaluate what an organization says – they also judge whether it understands how people feel.
Why Human Judgement Still Matters
Human-led crisis communications are critical to ensuring an organization has the best chance of withstanding the challenges a crisis presents, as well as recovering and rebuilding its reputation once the crisis has passed. Without someone looking at the crisis response through the lens of how an audience will receive the message, particularly those who were negatively impacted by the situation, organizations risk compounding the crisis and digging themselves into a hole they may not be able to escape.
Professional communicators aren’t just generating generic materials. We’re analyzing the full situation, looking at who was impacted, and working to intentionally put ourselves in their shoes. From there, we craft thoughtful and impactful communications designed to both protect an organization’s reputation and rebuild trust with its key constituents.
AI as a Tool, Not a Replacement
To be clear, I am by no means trying to wholly discredit AI or even suggest it shouldn’t be a tool that communicators use. That ship has sailed, and, like the internet and Microsoft Excel, some form of AI is here to stay. Used thoughtfully, generative AI and LLMs are most effective when they are applied to:
- Accelerating drafting and scenario planning
- Pressure-testing messaging frameworks
- Supporting, but never replacing, human decision-making
When paired with experienced communicators, these tools can help elevate the quality and speed of crisis response without sacrificing judgement or empathy.
The Bottom Line
When a crisis hits, empathy, judgement, and accountability aren’t optional, and they can’t be automated.
AI can draft, but humans must lead.