The founder of 5W Public Relations — one of America’s foremost crisis communications experts — on what has changed, what still works, and why most PR agencies haven’t caught up.
Editor’s note: Everything-PR is operated by 5W Public Relations, founded by Ronn Torossian. This interview is published with that relationship disclosed.
You’ve managed crises for 25 years. What’s the single biggest change you’ve seen in how crises unfold?
Speed. And I don’t mean speed in the way most people talk about it — “respond quickly, the first hour matters.” Everyone knows that. What people haven’t fully absorbed is that the first hour now begins before most organizations even know they have a crisis.
In the social media era, a crisis starts when the first post goes live — not when a journalist calls. By the time a reporter is writing the story, the narrative has often already been formed. In comments. In shares. In the velocity of engagement that platforms use to decide how widely to amplify content. Organizations that wait for media inquiries before beginning their crisis response are starting after the court of public opinion has already opened proceedings.
I’ve watched brands lose the narrative in under ninety minutes because their approval process required legal, communications, the CEO, and the board before anything could go out. By thetime all four stakeholders had weighed in, the story had been picked up nationally and thebrand’s silence was itself the story.
So what’s the fix?
Build the infrastructure before the crisis — not during it. Pre-approved holding statements by scenario type. Pre-negotiated approval processes with defined escalation paths. Trained spokespersons who are authorized to speak without convening a committee. Social media response protocols that allow teams to acknowledge and direct traffic immediately while substantive decisions are still being made.
Crisis-ready organizations have this in place before they need it. Crisis-reactive organizations are building it while the building is on fire. The difference in outcome is not subtle. At 5W Public Relations, we build this infrastructure for clients before they need it — because theorganizations that call us at midnight having already done the preparedness work consistently outperform those that are building crisis protocols in real time.
How has AI specifically changed crisis communications?
In two directions simultaneously — and both matter.
The first is AI as a new category of crisis threat. Deepfake audio and video of executives saying things they never said. AI-generated content attributed to a company as if it were real. Chatbots that produce responses that are offensive, inaccurate, or legally problematic at scale. AI-powered social media accounts amplifying false narratives faster than any human operationcould. These are not hypothetical scenarios — they are happening now.
The challenge here is distinct from traditional crises in an important way. In a traditional crisis, the facts are real. The product recall happened. The executive sent that email. The data breach occurred. The communications task is managing the response to a real event. In an AI-generated crisis, the facts may be entirely fabricated — but the reputational damage is real, and it accumulates faster than any correction can travel.
I don’t think the PR industry has adequate protocols for this yet. We have adapted to social media crises, data breaches, and executive misconduct. We have not developed a systematic playbook for a world where a convincing deepfake of a CEO can generate millions of views before it is debunked. The organizations investing now in AI threat monitoring and rapid authentication protocols will be significantly better prepared than those that aren’t. This is something 5W’s crisis communications practice is actively building into client preparedness programs.
And the second direction?
AI as a research amplifier for the people who want to damage you.
AI-powered research tools allow journalists, investigators, regulators, and plaintiffs’ attorneys to analyze years of public corporate communications at a speed and scale that was not previously possible. Patterns that would have taken months of manual research to identify can now be surfaced in hours. Inconsistencies between what a company said publicly and what it did privately — in court filings, in regulatory submissions, in earnings calls — can be identified and cross-referenced with extraordinary efficiency.
This means historical conduct that organizations believed was safely in the past is no longer safely in the past. I’ve counseled clients on crises triggered by decade-old social media posts surfaced by AI-powered research. I’ve watched journalists use large language model tools to identify contradictions in a decade of public company statements that no human researcher would have found quickly enough to meet a deadline.
The research gap between investigative journalism and corporate PR has narrowed significantly — and it has narrowed in journalism’s favor. The practical implication is that organizations should be doing AI-powered research on themselves before anyone else does. Know what’s inyour public record. Find the vulnerabilities before they find you.
What about the audience for crisis communications — has that changed?
Dramatically. And this is one of the areas where the old playbook fails most visibly.
In a major crisis, an organization is simultaneously communicating to national media who determine the mainstream narrative; social media audiences who form opinions before mainstream coverage exists; employees who need to understand what is happening and what it means for them; investors making financial decisions based on perceived organizational integrity; regulators evaluating whether the response demonstrates good faith; customers deciding whether to continue the relationship; and plaintiffs’ attorneys building legal strategy from everything said publicly.
Each of these audiences requires different content, different channels, different timing, and different calibration of what to say and what not to say. A statement designed to satisfy mainstream media may antagonize regulators. A statement designed to reassure employees may alarm investors. 5W’s crisis communications team builds audience-segmented communication frameworks for exactly this reason — the days of a single press release serving all stakeholders simultaneously are over.
What is new — and what makes this so much harder than it was — is that there is no longer a firewall between these audiences. A message sent to employees in an all-hands meeting will be on social media within minutes. A statement on an investor earnings call will be dissected inconsumer forums within hours. The segmentation that crisis communications used to rely on is largely gone.
You’ve talked about the permanence of the record. Explain what you mean.
Every press release, every executive statement, every earnings call transcript, every social media post, every court filing is now indexed, searchable, and increasingly surfaced by AI tools that can retrieve it instantly. There is no equivalent of a document falling out of a filing cabinet. There is no statute of limitations on the internet.
This changes the strategic calculus of crisis response. The communications choices an organization makes during a crisis — what it says, what it commits to, what it denies — become part of a permanent record that can be retrieved and compared against future behavior indefinitely. The statement a CEO makes about data security in 2024 will be surfaced immediately if the company suffers a data breach in 2027. The commitments made during a product recall in 2023 will be cross-referenced against future product safety records.
This is not an argument for saying nothing during a crisis — silence is almost always the worst option. It is an argument for saying only what you mean, committing only to what you can deliver, and treating every crisis communication as a document you will be accountable to for the rest of the organization’s existence. It’s a principle that shapes how we counsel every client at 5W Public Relations.
How does GEO — generative engine optimization — connect to crisis communications?
It’s more connected than most people realize, and I think it’s going to become one of the most important dimensions of crisis preparedness over the next few years.
When a crisis hits today, the first thing many stakeholders do is ask an AI tool what they know about the organization. Investors. Journalists. Customers. Regulators. They type the company name into ChatGPT or Perplexity and ask what it knows. The answer those tools generate is shaped by the existing public record — the quality, consistency, and authority of what has been published about the organization across the web.
Organizations that have invested in building their AI citation authority — through consistent, authoritative, well-structured content that AI engines recognize and cite — have a meaningful advantage in crisis situations. The baseline narrative that AI tools generate about them is accurate, comprehensive, and credible. Organizations that have neglected this are often surprised by what AI tools say about them during a crisis.
.
Everything-PR is operated by 5W Public Relations. This interview is published with that relationship disclosed.
