TL;DR
AI can summarize months of coverage in seconds, but it can’t tell you whether the data is sound or what your organization should do next. As information gets cheaper, judgment gets more valuable. The edge in 2027 belongs to the people who turn information into decisions.
Here’s the uncomfortable truth: AI is pretty good at a lot of what we used to call the craft. Drafting. Formatting. Matching tone. Research.
The result is a flood of competent but interchangeable output. Everyone can produce more. Fewer people can tell you what any of it means.
And when nobody’s checking? Earlier this year, I wrote about a professional friend who lost her job after a fabricated statistic turned up in a published article. The review that followed found several more AI-invented claims. Her work had been fast and polished. Nobody had verified it.
This is the third and final piece in our series on rethinking the PR tech stack. In part one, the CEO asked “So what?” and we talked about reporting meaning, not just math. In part two, we talked about buying tools against real questions instead of demos. Now for the most important part of your stack, the part that isn’t software: the person who answers.
Information Isn’t Intelligence
A platform can find thousands of mentions, flag an emerging topic, calculate share of voice and summarize a quarter of coverage before you finish your coffee. Great. Now what?
Getting information is no longer the hard part. The hard part is deciding which information you can trust and what it means. Someone still has to judge whether the underlying data is sound, whether the pattern matters and what the organization should do next. That’s the difference between information and intelligence.
It matters inside your own building, too. Comms teams earn trust with leadership by showing their work: where the evidence came from, what it does and doesn’t tell us, and why the recommendation makes sense in context. More data, delivered faster, doesn’t automatically make your advice more convincing. AI also makes it dangerously easy to produce a 40-page strategy document that says absolutely nothing. The tools will happily hand you volume. Meaning you have to supply yourself.
That gap between information and intelligence? That’s where your people live.
Pencil, Not Pen: From Data to Action
The path from raw data to strategic value runs in five steps:
- Information
- Context
- Insight
- Recommendation
- Action
Tools are strongest at the start. They collect information, organize huge datasets and increasingly help spot patterns. Human expertise matters more with every step toward action.
I think of AI as a pencil, not a pen. A pencil is for drafting and working through ideas, and you erase and refine as you go. A pen is for final decisions and published work. Too many teams are handing AI the pen.
What Good Analysts Actually Do
- Understand the limits of the data they’re working with.
- Connect signals from different sources.
- Separate meaningful change from noise.
- Know the business, the audience, the competition and the objectives.
- Translate all of it into a recommendation someone is willing to act on.
Some of the best signals are low-tech. A small circle of trusted field leaders you text after a big announcement will often tell you something is off before any dashboard does. People also see what the machine can’t. AI is trained on historical data, so it tends to reinforce existing patterns. The open angle, the story nobody else is telling, usually comes from a person.
AI doesn’t reduce the importance of human expertise. It increases it.
New Doesn’t Mean Ready
Autonomous research agents and synthetic audiences are coming, and their long-term potential is real. But they aren’t ready to make decisions you can’t verify, and most of our industry seems to agree: just 12% of PR pros use AI agents, and only 7% are comfortable letting them act on their behalf.
Want a cautionary tale? Look at AI-written pitches. More than 53% of journalists say they’re actively opposed to receiving them. Take the judgment out of a relationship and you lose the relationship.
So set the rules now. Decide which platforms are approved, what client information never goes into a tool, how output gets verified and who reviews it before anything goes out the door. And when someone asks where all of this is heading, give the honest answer. “I don’t know” is a power move when you follow it with what you do know, what you’re testing and when you’ll know more.
Invest in People Like You Invest in Platforms
When basic analysis is available to everyone, the differentiator becomes the quality of your questions, your interpretation and your judgment. That’s always been the real job. As I wrote about press releases, half of what agencies do is telling a client their news isn’t news yet, then helping them find the version that is. No tool does that.
Which raises a question I think about a lot: if AI handles more of the early-career drafting and digging, how do junior people build judgment? It won’t happen by osmosis. Make the apprenticeship deliberate. Walk through the why behind edits, let people make the call and talk it through afterward, and treat review sessions as teaching time, not just quality control.
So here’s your last assignment in this series: put your software budget and your training budget side by side. If one keeps growing and the other doesn’t, you know where your stack is weakest.
Go back to that silent room from part one, where the CEO asked “So what?” and nobody had an answer. The right metrics, the right tools and a well-run stack all lead to the same place: someone who can look at the information, say what it means and tell leadership what to do next. Invest in that person, and you’ll still have an answer in 2028.
This is part three of a three-part series on rebuilding PR measurement for 2027. Catch up on part one and part two. For more strategies on communication that drives results, listen to Everything Speaks.