5 Reasons You Shouldn’t Use AI to Write Your Crisis Communications Manual

In the age of AI-everything, it’s tempting to hand over complex tasks to algorithms – including the drafting of crisis communication manuals. But while AI might help you create informative blog posts for your organisation’s website (though not in this case!), a crisis comms manual is something else entirely.

These documents aren’t just about words. They’re about strategy, risk, ethics, and trust – none of which AI can reliably deliver (currently). Relying on AI to write your manual could not only be unwise – it could also damage your reputation and land you in legal trouble. 

Here are five reasons why the creation of crisis communications manuals is a job that still demands human expertise: 

1. AI Doesn’t Understand Legal Risk

Crisis communications must work in tandem with legal considerations. We’re talking about admissions of liability, the timing and wording of apologies, confidentiality obligations, and language that might have costly implications in legal proceedings or even triggering Contempt of Court if mishandled.

AI models aren’t trained to grasp these legal nuances. In fact, testing has shown that they often lack the judgment or regard for legal frameworks entirely. In a high-stakes scenario, for example a fatality, allegations of employee misconduct, or litigation, an AI-generated statement could do more than miss the mark. It could land you in court. 

2. AI Language Sounds Like… AI

There’s a recognisable tone to AI-generated content. It can be generic, overly informal, chatty and emotionally tone-deaf. In a harrowing crisis, using templated, robotic language doesn’t just fall flat – it can look insincere and even disrespectful. 

Worse still, it can provoke public backlash and cause a reputational crisis on top of the original one. People expect humanity in dark moments – not Chat-GPT-generated empathy. 

3. Writing a Manual Is a Strategic Process

Creating a crisis communications manual isn’t just about producing a document. It’s a form of management consultancy in itself. The process often uncovers gaps in internal knowledge, legal preparedness, insurance cover, or external support. 

AI can’t ask the right probing questions and follow-up questions. It won’t flag a missing insurance contact, or spot that you don’t have a legal process in place for internal investigations. A good crisis comms consultant will, and that insight is key to protecting communities and reputations. 

4. Training and Testing Require Human Experts

Once a manual is written, a key part of gauging its effectiveness lies in stress testing it. Who understands it? Who owns it? Can your team apply it under pressure?

Scenario training, Q&A sessions, and real-time simulations are essential – and they work best with experienced crisis professionals who can adapt to your organisation and provide live feedback, not canned answers. 

AI can’t read the room. It can’t see confusion in your team’s eyes, sense when minds are becoming overloaded or pivot when a training session reveals a fundamental misunderstanding. 

5. Manuals Need to Evolve – and Experts Know When That’s Needed 

Crisis risks aren’t static. A good advisor will proactively recommend updates as new threats emerge – from deepfakes and fake journalists, to AI-generated blackmail or cyberattacks.  

AI can’t track evolving trends or interpret how they apply to your business. A seasoned expert can – and they’ll often prompt updates before you even realise they’re needed.  

Final Thought: Use AI as a Tool, Not the Author

AI might help you format documents or brainstorm talking points. But when it comes to writing your crisis communications manual, you need human expertise, not automation. 

If you’re building a manual that could one day defend your brand, protect your people, and reassure the public – make sure it’s written by someone who understands what’s really at stake.