AI Prompts for Crisis Communication

Picture this: you're a comms director at a mid-size consumer brand. It's 7:43 AM on a Tuesday and your phone is already full of Slack notifications. A supplier just got named in a labor violation story—and your brand is mentioned in paragraph four. The journalist emailed at 6 AM asking for comment by 10.

You have two hours and a blank document.

This is exactly the scenario where ai prompts crisis communication workflows earn their keep. Not because AI writes your crisis response for you, but because it gets you to a structured draft fast enough that lawyers and leadership have something real to react to.

The Problem the Communications Director Faced

In a real crisis, the bottleneck almost never is strategy. Most experienced comms people know roughly what they need to say. The bottleneck is getting words on paper fast enough to get approvals in and still hit the deadline.

Without a structured prompting approach, teams waste the first 30-40 minutes arguing over where to start. With the wrong prompt, AI generates corporate boilerplate that can actually make coverage worse—the kind of non-apology statement that journalists screenshot and mock on social media.

The cost of that mistake isn't just embarrassment. It's the follow-up story: "Brand Fails to Address Concerns." That story didn't have to exist.

The Wrong Approach

The most common mistake: "Write a crisis statement for our company about [situation]."

The output is predictably bad. It sounds like every other corporate non-response you've read: "We take these matters very seriously and are committed to [generic value]." That kind of statement doesn't help you. It signals to journalists—and the public—that nobody at the company actually engaged with the situation.

⚠️ Common mistake: Treating crisis communication as a writing problem. It's actually a decision problem. Before you write anything, you need to decide three things: What do we know for certain? What do we not yet know? What are we committing to do? AI can help you structure those decisions—but only if you prompt it to do that work first, not skip straight to drafting.

The Correct Prompt

Here's a crisis communication prompt that forces the right decisions before any writing starts:

Act as a crisis communications consultant with 20 years of agency experience, including corporate litigation and product recalls. I'm dealing with this situation: [describe in 3-4 sentences, clearly separating what is confirmed from what is still unclear or alleged].

Build me a crisis response framework before I draft anything public:
1. Draft a holding statement under 75 words that acknowledges the situation without admitting fault for anything still unconfirmed
2. List 5 specific phrases or claims we must NOT make in any statement right now, with a brief reason for each
3. Rank the stakeholder groups we need to reach in the next 24 hours by urgency: employees, media, customers, partners, investors, regulators
4. Suggest one proactive action we could announce that demonstrates accountability without conceding liability
5. Flag any element of this situation that could escalate significantly if not addressed in the next 6 hours

Be direct. I'm under deadline.

What this does: The "must NOT say" list is the element most teams skip. It forces the AI—and your comms team—to think defensively before putting anything in writing. In practice, this list catches more legal exposure than any other part of the prompt. I've seen teams avoid serious problems just from this step.

⚡ Pro tip: Run this prompt before you write a single public-facing word. The stakeholder ranking alone routinely surfaces audiences—employees, retail partners, regulatory bodies—that comms teams forget about when they're focused on managing the media story.

Results and What Changed

A corporate affairs manager at a food manufacturing company used this framework when a product recall story broke without warning. The prompt generated a holding statement in under 2 minutes. The "do not say" list immediately flagged two phrases that were already in the first instinctive draft—both of which legal confirmed were problematic.

Even more importantly, the stakeholder mapping reminded the team that retail partners needed direct outreach before the public statement went out. That call to major retail accounts happened 40 minutes before the press release—and it protected several key relationships that would have been blindsided otherwise.

Total time from blank screen to approved holding statement: under an hour. Previous crisis protocol without AI: half a day of circular discussion before a word got written.

Prompts for Specific Crisis Types

Social media backlash:

Our brand is trending on [Platform] because [describe situation]. Current sentiment: [negative/mixed]. Draft a 2-post response thread that: (1) acknowledges the specific concern without being vague, (2) states concretely what action we're taking, (3) invites DMs for individual situations. Tone: direct and human, not defensive or performative. Do not use the phrase "we hear you" or any variation of it.

Employee communication:

Write an internal memo from [CEO/SVP] to all employees about [situation]. Employees are likely already seeing external coverage. Be honest about: what we know, what we don't know yet, and the specific next steps the company is taking. Keep it under 300 words. No corporate platitudes—employees will see through them immediately and it will make trust worse, not better.

Media inquiry response:

A journalist from [Publication] is asking about [specific allegation or situation]. Confirmed facts we can speak to: [list]. Facts we cannot yet confirm: [list]. Draft two separate items: (1) a 3-sentence on-record statement, and (2) talking points for a phone call including one bridging phrase to use when asked about something outside our confirmed facts.

What this does: Separating on-record statements from verbal talking points is a distinction AI prompts almost never make on their own. Your written statement is permanent record. Your talking points give your spokesperson flexibility without going off-message—they serve completely different functions.

⚡ Pro tip: After any crisis response goes public, immediately run this: "Analyze this statement we just published. What questions does it leave unanswered that a reporter covering this story would definitely ask in the next call?" The gaps it identifies are exactly what your spokesperson should prep for before their next interview.

How to Apply This to Your Situation

The specific crisis type matters less than the framework. Whether it's a data breach, an executive misconduct allegation, a supply chain failure, or a product defect—the prompting sequence is the same:

1. Brief the AI on confirmed vs. unconfirmed facts as separate inputs
2. Get the "do not say" list before anyone drafts anything
3. Map stakeholders before you start producing multiple versions of communication
4. Draft each piece for its specific audience—not a generic "public statement"

⚡ Pro tip: Build a "company context document" for crisis situations now, before you need it: confirmed company values language, current leadership roster with titles, standard legal disclaimers for your industry, regulatory body names and contact information. Paste this into every crisis prompt as context. It stops the AI from inventing specifics that then have to be corrected under deadline pressure.

The companies that handle crises well aren't faster writers. They're faster decision-makers. These prompts accelerate the decisions, not just the drafts.

Next Steps

Build your crisis communication prompt library before you need it. Seriously—this is one of those areas where preparation has an outsized return. The worst possible time to figure out your prompting workflow is at 7:43 AM with a journalist deadline in two hours. PromptABCD lets you save these frameworks, tag them by crisis category (reputational, product, personnel, regulatory), and share them across your comms team so everyone pulls from the same tested prompts when the situation is live and the clock is running.

After the Crisis: The Debrief Prompt

One thing almost no comms team does: run a structured debrief after a crisis to capture what worked and what to do differently. Most teams are just relieved it's over. But the institutional knowledge from handling a real crisis situation is genuinely valuable—and it evaporates fast.

We recently handled a communications crisis about [topic]. Here's a brief summary of what happened: [describe the situation, the timeline, and the communications we put out]. 

Conduct a post-crisis debrief analysis:
1. What did we communicate that worked well—and what specifically made it effective?
2. What did we communicate too slowly, and what was the cost of the delay?
3. What stakeholder group felt most informed? Which felt least informed?
4. What should our holding statement template look like for a similar situation in the future?
5. What one policy, template, or protocol would have made this significantly easier to manage?

⚡ Pro tip: Run this debrief prompt within 72 hours of the situation resolving, while the details and decisions are still fresh in everyone's minds. The output becomes a real training document for your team and a concrete baseline for improving your crisis communication protocols over time. Save it in PromptABCD alongside your crisis prompt templates so future comms hires can learn from real situations rather than hypotheticals.