In this episode of The PR Breakdown, crisis communications consultant Molly McPherson explains how the rise of social media and AI has fundamentally altered crisis management. She argues that the old buffer of time—used to assess, shape, and control a narrative—has vanished. AI now aggregates and summarizes social‑media chatter in real time, creating a de‑facto “first draft” of a story before any organization can respond. Waiting, even briefly, hands the narrative to AI, which presents a compiled version of public sentiment as truth. McPherson stresses that the biggest risk is not saying the wrong thing, but being summarized at scale. She advises leaders to prioritize clarity and trust over perfect information, using rapid, honest statements that acknowledge what is known, unknown, and the next steps. Trust is presented as the core metric of a crisis; its erosion can be detected through behavioral signals—tone, silence, who is speaking—rather than solely through dashboard metrics. She illustrates the point with a real‑time case study of Delta’s CEO Ed Bastian responding to TSA staffing issues that stem from a political shutdown, showing how proximity to stakeholder pain, not fault, defines a crisis. The episode concludes that human leadership, instinct, and judgment must work in harmony with data and technology to navigate modern AI‑driven crises.
Chapters
Let me tell you what's quietly changed in crisis communication, and most organizations and their leaders still haven't caught up to it yet. I'm someone who works with a lot of different clients in a lot of different sectors. I also work with online content creators, people who rely on their reputation. And it's not just that things are faster, it's that you don't get time anymore to set your reputation. Time used to be your buffer. Time used to be your strategy. You could pause, you could assess, you could shape the narrative before anyone else did. Think about it. Back before social media, if there was a story there, a fixer could be representing someone, it could be a lawyer, It could be someone from the company. All they had to do is make a phone call to an editor at a newspaper or make a phone call to a network. Maybe it was an agent in Hollywood with a celebrity. Maybe it was political. There could be some bartering. Now reputation becomes transactional. If you kill this story, we'll give you this story. But the days of being able to control the narrative human to human, That's over. Hey there. Welcome to The PR Breakdown. I'm your host, Molly McPherson, someone who works on the front lines of crisis. I am with clients or I'm watching it on the outside just like you. And I want to share what I'm seeing right now. Here's the shift. First, social media changed the game. You all know that. The public started shaping the story in real time. Organizations, they don't love it. Leaders, people at the center of it, they don't like it at all. But a lot of them, I notice, still dismiss it. You'll see this in big celebrity court cases when they have their bulldog lawyers or a response from a celebrity, any celebrity where they blame cancel culture, they blame the public. Any response like that, I say that person simply doesn't get it because they're dismissing it as noise. In my work, I have clients who dismiss social media as noise. But now, now we're in an AI-driven environment, and AI doesn't ignore that noise. It organizes it, it summarizes it, and it presents it as truth before you even respond. So here's the real shift. You are no longer the first draft of that story. If you wait in a crisis, even for a short time, a version of the truth gets built without you. So good luck trying to undo it. The biggest risk right now isn't saying the wrong thing. It's being summarized at scale. Think about that. A customer, someone is searching your company or searching your name, or a reporter is using AI, Perplexity, Claude, Gemini to get up to speed on a story, getting up to speed on your story. You could be a CEO and your board is asking, what is going on? And they go to AI. AI isn't just putting out someone's statement. If someone's in a crisis, ChatGPT doesn't just put out your statement. No, the person searching is getting a compiled version of everything that people are saying about them. Reddit, TikTok, news, everything. And if you haven't shown up yet with a response, that version becomes the default. This is where I see companies make the same mistake over and over again. Celebrities, politicians, they dismiss the early signals. They say, it's just social media. It's just noise. It's a few idiots complaining. I hear that one a lot. Uh, we don't have all the facts yet. Just let it be. But that noise, that's exactly what AI is using to define your reputation. So by ignoring it, You are not avoiding a crisis, you're cementing it. And if you haven't been able to tell yet, this episode is for the communicator, the people who are on both ends. We see the tech, we see the inside, we see the outside, we know the culture, we know what's happening inside the building, but we also know what people are saying about the building. Add to that speed. Let's talk about speed because people always ask, How do you balance speed with accuracy, especially in this AI age? Here's the truth, at least from my perspective. Waiting does not protect accuracy anymore. Waiting surrenders the narrative. Before, it was to the press doing a story, or it's to social media. But now AI, it's like that. You can't wait. Now, you don't need perfect information to respond. You just need clarity. Clarity. And those AI-written statements, they cannot seem like they come straight from AI. That doesn't work. You want to talk about, in a crisis, something goes on. What do we know? What don't we know? What are we doing next? The goal is to build trust. Write a statement that builds trust because reputation now is not built on these statements. So don't sweat it out. Don't make ChatGPT write every single word. Trust is built on signals. It's built on volume, velocity, credibility mixed with emotion. Who's talking about you? How fast is that information spreading and how do people feel about it? Because that's what AI is pulling from, which means this, you, even you personally or your organization, you are training AI on your reputation every single day. Whether you realize it or not. Just go ahead, Google yourself, see what comes up with SEO, but then go to Perplexity and see what Perplexity has to say about you. See where it polls, particularly if you're online. Are you a content creator? Are you a leader? AI is going to give you a better assessment of who you are and where you rate from a reputation point of view than a simple Google search. And here's something else leaders don't like to hear, leaders who I work with. You don't have to cause a crisis to be in one. If your customers are frustrated, if your stakeholders are impacted, even if you didn't cause it, you're in it. A crisis today isn't only defined by fault, though in many cases it is fault, but sometimes it's someone else's fault and it's defined by your proximity to that pain, to the stakeholder's pain. So how do you know if trust is actually breaking? Trust, it's my metric for if you're in a crisis or not. If you're losing it rapidly, you're in crisis. You don't do it by looking at a dashboard first. It doesn't mean going to your social media dashboard, like your Sprout Social or something, to see what people are saying. If you're a company who has a media database access. You could be Cision. I work with Muck Rack. I can see all the data points. If you rely too heavily on that, you're gonna be in trouble. You wanna look at what the dashboards are saying, what the metrics are saying, but you also wanna measure it with behavior. Look for tone, look for silence. Who's speaking? Who's not speaking? Does it match what other people are expecting? Customer expectation is huge. Managing that expectation is huge. Trust doesn't break all at once. Trust, and I know this from working with people in crisis, it leaks, and it usually leaks from within, like the building, like literally the four walls or the virtual walls. If you don't have good culture, that's gonna be a place that's gonna be ripe for crisis or a trust crisis, and then it just spreads out of the building. And if you are not paying attention and it's your building, that leak spreads fast. So how do you know if trust is actually breaking? This is how I explain it to leaders. I make it very simple. If trust is breaking, revenue is next. Customers disengage, audiences organize, and then the media follows. And what starts as a perception issue turns into a massive operation risk or a legal risk or a financial risk. And it doesn't matter if you're the CEO a company or you're an influencer, dominoes fall and they fall quickly. So if you're someone thinking about AI and crisis management, crisis communication, don't just start with tools. Don't just look at the back end of TikTok or the back end of Facebook. Don't just look at your dashboards. Start with the truth. Start with trust. AI is going to surface the problem whether you participate or not. It's going to organize perception out there. That's how you look for sentiment. It summarizes sentiment. It will define the version of the story and put that out into the public. But how do you respond to that? It's important to remember that you need a human element in any type of crisis response because leaders are human. Leadership is a human process. And that's the part that no technology can replace because AI is going to surface a problem whether you participated or not. That's still leadership. And here's an example from, and here's a recent example that's playing out in real time. Look at what's happening in air travel. I am getting on a flight soon and I am so nervous. I do not know, do I leave 5 hours in advance or 2 hours in advance? The TSA staffing issues, the long lines, people missing flights. I was talking to my friend Debbie last night. We were talking about Houston Airport. She was flying out of Texas, hours and hours and hours. That frustration doesn't just contain itself to one agency, TSA. It spilled to the administration because it's a shutdown. TSA did not cause the problem. There is a shutdown. There is a block. That block is called President Donald Trump. It is a Washington, D.C. problem that is spilling out and affecting the real people, the average people, not the billionaires in their private jets, but the regular middle-class people, people like me getting on a flight soon, getting on planes, long lines, people missing flights. That frustration now bleeds into the airlines. It bleeds into the airports. Delta is an example. They did not cause the problem. They're doing whatever they can to keep up with it, but their customers are still the ones standing in line for hours. Well, this is all the airlines, but Delta is an example of someone stepping in. First, the CEO, Ed Bastian, joined other airline leaders in coming out and calling on Congress to fix the issue and ensure that TSA workers are paid. Getting ICE agents to help TSA agents is not solving the problem.
It's inexcusable that our security agents, frontline agents that are central to what we do, are not being paid. And it's ridiculous to see them being used as political chips. So, you know, we're outraged, you know, and if there's a call to action here, and I think over 90% of the American public support those people getting paid, ask our folks right here in Washington, to do their job, get our people paid. They can do it.
But the shift here is that Delta didn't just say, hey, this isn't our fault. Don't be mad at us. That's a response that we see a lot. They said, this is affecting our passengers, our people, so we're going to show up. Because in today's environment, crises, they're not defined by faults all the time. They're defined by the proximity to the fault. And the proximity to the pain, people's pain. AI captures all of that, the frustration, the delays, the response. Imagine people on their phones. Imagine how many times the word Delta, you know, comes up in that frustration. But it also comes up with TSA, White House politics, blah, blah, blah. But what people remember in the end is who caused the pain and who showed up for them to ease it. And that's the part no technology can replace. That's all for this week on the podcast. I wanna leave you with this. When you are watching a crisis or you're trying to prevent one in your own company or even for yourself, don't immediately rely on technology and AI and dashboards. Watch the signals, trust your instincts, then go to the data. See what it says, and put them all together. Human and data and technology working in harmony. That's how you get through a crisis. All right. That's all for this week. Bye for now.
