What AI Can’t Replace: The Future of Strategic Communications

Joyce Lee is an Account Director at Montieth SPRG, a boutique global PR and strategic communications firm with teams in Hong Kong, New York, and London. With over 13 years of experience spanning agency and in‑house roles, advising clients across financial services, technology and corporate affairs, Joyce writes from the front lines of an industry being reshaped by AI. In this piece, she examines what AI can and cannot do, and why the fundamentals of great communications have never mattered more.  

There’s a question I keep coming back to lately: who is actually behind the words we read?  

Does it really matter? 

The truth is, it is getting a lot harder to tell. What you are reading could come from a seasoned industry veteran, someone fresh out of university, or a model trained on half the internet. Sometimes, the quality gap is not always obvious. It starts to feel less like a voice and more like a template.  

How Communicators Learn and What AI is Changing  

One of my first jobs as an intern was to draft a press release. I came from a hotel management background, so as a novice, it felt intimidating at first. I still remember how I stared at a blank white screen for hours, poring over dozens of past client releases, trying to find the right tone and voice. 

What followed was one of the best lessons I didn’t know I needed. After submitting my draft, my manager would print it out and mark it up line by line. Red pen everywhere: slashes, arrows and rewrites in the margins. It reminded me of the first day of school. She would explain out loud why this word didn’t work; why that sentence lacked clarity.  

It was uncomfortable and honestly a little painful, but it helped me learn how to do it a bit better the next day, every single day. I thought of her when I read On Writing: A Memoir of the Craft by Stephen King, where he describes writing as a craft passed down through practice, critique, and the quiet generosity of those willing to teach. And that’s where things start to feel uncertain now. 

Much of that process – the messy, repetitive, early-stage work – is exactly what AI gets very good at once it’s trained: drafting, refining, iterating. Those were the kinds of tasks that used to be how you learned. Now, that learning layer is being stripped away for those just entering the market. 

The AI Content Flood and the Premium on Credibility 

AI models have already created a flood of verbose, generic content. And we are slowly witnessing brand leaders rediscovering that what still differentiates isn’t volume, but the ability to frame a story that readers, regulators, investors and machines recognize as credible. 

The double-edged truth here is: agencies cull their juniors in a relentless quest to maintain margins, yet the market is paying a premium for communicators who can set and defend the narrative. This means being able to create content that humans and machines can interrogate and trust, as well as understanding how AI models read and cite content – and it’s not random. 

The data backs this up. According to Muck Rack’s latest Generative Pulse report, press releases cited by AI contain around 30% more objective sentences on average than non‑cited content and most AI citations for a given brand come from only 20 outlets. The report also shows earned media sources account for nearly 25% of all citations generated by large language models. 

GEO Strategy: How to Optimize for AI Citation and Discovery  

Yet at the same time, AI models are constantly tuning their citation mix. Communicators can try to understand, but in reality, we are operating in a black box. Sometimes these citation trends change overnight. Sometimes it’s a slow adjustment. What surfaces today may disappear tomorrow, not because the content changed, but because the model’s citation logic did.  

This means communicators can’t just publish and forget. It is not a one‑off exercise. You now have to check how content shows up across platforms like ChatGPT and Claude on an ongoing basis. Brands need to actively test how their content is retrieved, cited, or ignored — running live searches, mapping citation behavior, optimizing accordingly, and then repeating the process months later to see what’s changed. This demands a precise, high-cadence earned program aimed at the specific journalists, wires and platforms that AI trusts for your category. 

AI is optimized for explanation, framing and recommendation, not just discovery. Design “explanation layers”: clear, consistent narratives about what you do, who you serve and why you’re credible that hold up across thousands of prompts. 

The Human Judgment in PR  

I don’t think there’s any need to romanticize the old days. A lot of junior work was repetitive, sometimes mind-numbing. AI taking that off people’s plates? Probably a good thing. 

While AI is prolific, it doesn’t know when something crosses the line. It can’t reliably judge when a claim overreaches. A narrative can be technically accurate and still misleading, and AI has no instinct for where that line sits or when it has been crossed.  

Going back to the basics, marketing communications and PR have always been about protecting reputation and building trust, especially when things get difficult. As Warren Buffett famously said, “ It takes roughly 20 years to build a reputation and five minutes to ruin it.” People can tell when messaging feels generic or recycled, and that becomes very obvious in a crisis. A templated response might tick the right boxes, but it can come across as distant or performative. This is where human judgment becomes non-substitutable. 

Reading the room, picking up on nuance, and responding in a way that genuinely reflects a situation is an art. Sometimes that means being more specific, a bit more candid, or simply sounding like a real person. You can’t really rely on AI to find this balance – it comes from experience, context, and an understanding of not just what people are reacting to, but why. 

What This Means for Communicators and Organizations  

From generation to selection, from volume to credibility, the role of PR is shifting. But the fundamentals in PR are still the same: why this matters, what’s happening, how to explain it in a way that holds up. What has changed is everything around it. Now those fundamentals need to be accompanied by data, critical thinking, and a willingness to reject narratives that are elegant but flawed.