Harnessing the Power of AI for PR: The Do’s and Don’ts for Communicators

AI is showing up in PR workflows more than ever. But using it well? That takes nuance. In our recent blog, Did a Robot Write This?, we unpacked the growing presence of AI in creative work—and why authenticity still matters. Now we’re digging deeper into the practical side: how communicators can use AI as a tool without losing the human touch. From Harry’s Communications team, here are the do’s, don’ts, and best practices we’ve learned from experience.

AI Can’t Replace a Smart Pitch Derived from a Great Interview

When we’re pitching a story or responding to a media request for comment, every word matters. The goal isn’t just to get a quote placed—it’s to build credibility, offer real insight, and make the reporter’s job easier.

At Harry, our secret sauce is simple: we ask the right questions. We know how to unlock our clients’ expertise and turn it into media pitches that are sharp, insightful, and media ready. That’s why we always lead with conversation—interviewing clients, pulling out the gold, and crafting responses that reflect their voice and unique point of view.

But every now and then, schedules don’t align. A client sends notes after hours or while traveling. No problem—we take those notes and shape them into compelling pitches in their voice. What doesn’t work? When someone skips the process altogether and feeds a media request into AI.

One client did just that—used ChatGPT to draft a media response and sent it to us. We flagged it immediately. The formatting gave it away. The tone was off. And it just didn’t sound like them.

In another instance, we did conduct a full interview, crafted responses ourselves, and included a several industry phrases our client used in conversation. The reporter came back asking if AI had written it—turns out, those common phrases had been used in other articles. We shared raw notes from our meeting to prove the content was human and original. (Happy ending: we got the placement).

Lesson learned? 

We’re in a new era. Reporters are on high alert, and they’re rightfully wary of generic, AI-sounding responses. Some even use detection tools to sort the real from the recycled.

In this landscape, authenticity is everything. Reporters want expert insight—real voice, real value. And that’s where we come in. We don’t just send quotes—we build trust, both with our clients and the journalists they’re trying to reach.

Making AI Work for You: 6 Smart Tips from Team Harry

AI can be a powerful tool in public relations if you know how to use it. We use AI where it adds value: research, fact-gathering, refining ideas and adjusting tone. But strategy, voice, and credibility? That’s all human.

Here’s how we approach it:

1. Keep it human. Use storytelling, real anecdotes, and a conversational tone. Reporters don’t just want a quote—they want a source they can trust. Authenticity earns placements and relationships. The goal isn’t just this story. It’s the next one, too.

   
   

2. Source everything: Always ask AI to cite its sources. Click through to those links to find additional information and context. Sometimes, you may find outdated information that needs fact-checking. Don’t be afraid to challenge your AI assistant!

   
   

3. Fact-check: AI doesn’t always know the difference between yesterday and five years ago. So you have to ask it to check stats, dates, and names to ensure you’re using accurate and current information. AI is, after all, recycling what’s already online—which can mean old data, outdated ideas, and an unoriginal voice.

   
   

4. Don’t take no for an answer: If AI says it doesn’t know something—or worse, makes something up—refine your prompt. Get specific. Narrow the scope. Push it until it delivers.

   
   

5. Use it for advanced reasoning. If you’re swimming in reports, analytics, search results, surveys, and social media insights, AI can help you connect the dots. Tools like Perplexity, Chat GPT, Grok, and Gemini offer “deep research” modes that can help you summarize, spot trends and patterns, and synthesize information.

   
   

6. Protect your data: Don’t drop proprietary or sensitive information into public AI tools. Use private mode, scrub identifying information, and stay up-to-date on privacy and cyber safety practices.

 

AI is a Helpful Tool, Not a Replacement for Creative Strategy 

Clients hire us for our skills, judgement, strategic thinking, extensive experience—not for producing auto-generated content. That’s why we don’t rely on AI to write press releases, blogs, video scripts, op-eds, social media, or ad copy from scratch. What do we use it for? Working smarter.

We think of AI as our research assistant—not our writer:

• It’s fast and saves time on background research and fact-gathering

• It helps us get up to speed on new subject matter before we begin a project.

• It acts as a friendly editor—great for tightening a sentence, testing tone, or generating alternate headlines.

One member of our team puts it like this:

“In my first draft, I over-write. I’ll use AI to help me edit. But I never input a document in full. If you do, it comes back sounding canned. To get the best results, I submit short copy blocks to check grammar and spelling or to help improve sentence structure.”

AI can also fine-tune your content for specific audiences, helping you to home in on the right tone, making content punchier, softer, more attention-getting, or more aligned with industry jargon.

Our team favorites? ChatGPT takes the top spot, with Perplexity a strong second.

And in case you’re wondering—did AI write this blog? Only one line: the headline trim. The rest is 100% human, built on real insight from the Harry comms team.

Because when it comes to smart content, we know where to draw the (very fine, very strategic) line. Carolyn Mandelker is a communications director at Harry.