Who Trains the Next Generation of PR Pros in the AI Era?
I started in PR when every media contact lived on a shelf.
Not in a CRM. Not inside a searchable database with filters and tags. On a shelf, in thick reference books called Bacon’s Media Directories, now part of Cision’s media database. If you wanted to find the right reporter, you pulled down a volume, looked up the outlet, checked the beat, checked the title, then cross-referenced that with recent bylines.
Press releases went out by fax.
Coverage was proven with scissors.
You cut a placement out of the newspaper or trade publication, pasted it onto blank paper, and photocopied it. That is why we still call them “clippings,” even though no one has had newsprint on their hands from a client placement in years.
Pitching was just as physical. You picked up the phone and called a reporter cold.
“Hello, Walt Mossberg.”
“Hi Walt, it’s Joel with [client].”
“Who?”
“Joel. With [client].”
“What do you want, Joel?”
“I have a column idea. Do you have 30 seconds?”
“Make it quick.”
“Super. You’ve heard of text messaging, right?”
That was the job. You learned by doing work that was slow, manual, and uncomfortable. You learned how reporters thought. You learned what they ignored. You learned what happened when a pitch was too broad, too late, too self-serving, or just plain boring.
Fast forward a few decades. Tonight, I built a case study in about a minute. That same job once took an afternoon of research, writing, editing, design, proofing, and revision.
That is convenient. It is also a problem.
Because the people who used to do that work were early-career PR pros. The jobs were not glamorous, but they taught the craft. Now a large share of that work is automated, instant, and cheap. The output still exists. The training reps do not.
The old grunt work was really training in disguise
A lot of early PR work looked administrative from the outside.
Build a media list. Clip a story. Draft a case study. Find three reporters who covered a certain issue. Call a newsroom to confirm a beat. Read a stack of trades and flag anything relevant to a client.
It was repetitive. It was often tedious. It was also how people learned.
When you searched a media directory by hand, you had to think. The work forced questions that still matter:
Does this reporter actually cover this topic?
Have they written about this company’s category recently?
Do they quote vendors, customers, analysts, academics, or all of the above?
Is this a news story, a trend story, a product story, or a feature?
What would make this useful to the reporter’s audience?
A junior person who built 50 media lists the hard way started to see patterns. They noticed that a “technology reporter” at one outlet did not cover the same kind of technology as a “technology reporter” at another. They learned that a person’s beat was not the same as their interest. They learned that one reporter loved data, another needed customer examples, and another never wrote about private companies.
The same was true of clippings.
Every Thursday, when the trades arrived, you were either a hero or a zero.
That pressure taught people something no workflow tool can fully teach: earned media is not a deliverable you control. It is a result you earn.
AI did not kill PR, but it changed the apprenticeship
The profession is not shrinking because the work stopped mattering. If anything, earned visibility matters more now.
AI search leans heavily on credible third-party sources. Muck Rack’s May study found that 84% of AI citations came from earned media sources, while paid content accounted for 0.3%. That is a sharp reminder that press coverage does more than impress executives or fill a monthly report. It helps shape what machines retrieve, summarize, and present as knowledge.
Brands with broad press footprints can surface faster and more often in Perplexity, ChatGPT, Gemini, Claude, and other AI systems. Not every answer comes from media coverage, of course. But strong earned media creates public evidence. AI systems rely on public evidence.
That should be good news for PR.
It also raises a harder question: if earned media is becoming more important, who is learning how to earn it?
Because the old path had a built-in sequence.
You started with research, lists, clippings, drafts, and reporting. You absorbed the rhythm of news. You learned the difference between a client’s internal priority and a reporter’s actual story. Then you graduated to pitching, strategy, messaging, and counsel.
AI compresses that sequence. A junior person can now generate a media list, draft a pitch, create a case study, summarize coverage, and build a report in minutes. The work gets done, but the person may not understand why one answer is good and another is useless.
That is the new risk.
Not that AI will replace the entire PR function. The real risk is that AI will remove the friction that trained judgment.
The missing skill is not typing, it is judgment
A case study built in one minute can be perfectly readable and still strategically weak.
It may miss the real customer tension. It may bury the result. It may sound like every other case study. It may use the wrong proof point. It may say “transformation” three times and never explain what changed.
The same is true of AI-generated pitches.
A tool can produce 20 subject lines. It can summarize a reporter’s archive. It can draft a note in a friendly tone. But it cannot know, without human direction, whether the pitch is worth sending.
That judgment comes from exposure. It comes from reading enough real stories to know what makes a story real. It comes from getting ignored. It comes from hearing a reporter say, “That is not a story.” It comes from having to defend why a certain outlet belongs on a list and why another does not.
Early-career PR pros still need those reps. They just will not get them accidentally anymore.
The hard part is that the industry used to outsource training to inefficiency. Manual work slowed people down long enough to make them think. Now the work is fast by default. Speed feels like progress, but speed can hide shallow understanding.
A young pro can now produce acceptable-looking work without ever building the mental model behind it.
That is dangerous for agencies. It is dangerous for in-house teams. It is dangerous for clients who assume a polished asset means sound strategy.
We need to rebuild the training layer on purpose
If the old apprenticeship disappeared, the answer is not nostalgia. No one needs to bring back fax machines or clipping books. The answer is to design a better training system around the work as it exists now.
That starts with treating AI as a tool inside an apprenticeship, not as the apprenticeship itself.
A useful training model might include a few basic habits.
Make research visible again
Do not let junior staff hand over a media list without explaining how they built it.
Ask them to show:
What they searched
Which reporters they ruled out
Which stories informed the choice
What angle each reporter might actually care about
What evidence the client can offer
The goal is not to punish mistakes. The goal is to make thinking visible. A media list should not be a spreadsheet of names. It should be a map of relevance.
Require the human read
AI can summarize a reporter’s work. That does not replace reading it.
Give junior staff a rule: before pitching a reporter, read at least three recent pieces and write three sentences about the pattern. What does the reporter return to? Who do they quote? What counts as evidence in their work?
That small exercise trains attention. It also prevents lazy personalization, the kind that opens with “I loved your recent article” and then proves the sender barely understood it.
Turn AI output into editing practice
AI drafts are useful training material if teams treat them as drafts.
Ask early-career pros to mark up an AI pitch and explain what they changed. Cut the generic claims. Replace vague language with proof. Move the news higher. Remove the client-first framing. Add the reporter’s audience.
Editing teaches taste. It also teaches restraint.
A young PR pro who learns to improve machine-generated copy will often become a stronger writer than one who starts from a blank page every time. But only if someone reviews the edits with them.
Bring back the pitch lab
Cold calling reporters is far less common now, and many reporters prefer email or other channels. That does not mean pitch practice should vanish.
Teams should run pitch labs.
One person plays the reporter. One person pitches. The “reporter” pushes back.
Why now?
Why this outlet?
Who else is saying this?
What data do you have?
Can I talk to a customer?
Why should my readers care?
This kind of practice is uncomfortable. That is the point. It helps people learn how thin a pitch can sound when it leaves the safety of a document.
Managers have to stop confusing output with growth
AI makes it easier to produce more work with fewer people. That creates a tempting management fantasy: if the tool can do the junior work, maybe the team needs fewer juniors.
That logic may work for one budget cycle. It fails as a talent strategy.
If no one enters at the bottom, no one matures into the middle. If no one learns the mechanics, no one develops the instincts. If every entry-level task gets automated away without a replacement training path, the industry will eventually wonder why it has plenty of operators and not enough counselors.
Senior PR judgment does not appear fully formed. It is built through years of small corrections.
A senior person knows when the client does not have news. They know when a spokesperson quote says nothing. They know when a reporter’s silence means “no,” when it means “not yet,” and when it means “you buried the lead.” They know when to push back on a client who wants coverage without substance.
That wisdom comes from pattern recognition. Pattern recognition comes from reps.
If AI removes the reps, managers must create new ones.
The next generation needs harder assignments, not busier calendars
Training does not mean giving junior people fake work. It means giving them real work with real feedback.
A strong early-career PR training plan should include assignments like these:
Build a media list, then defend every name on it.
Read a reporter’s last five stories and identify the common thread.
Rewrite an AI-generated pitch so it becomes specific, timely, and credible.
Draft a spokesperson quote that sounds like a human being.
Analyze a piece of coverage and explain why the client did or did not matter in it.
Find the missing proof point in a case study.
Sit in on a client call and write down what is actually newsworthy, not just what the client wants announced.
Practice saying, “I do not think this is ready to pitch yet,” with a clear reason.
That last one matters.
The profession needs people who can produce assets. It needs people who can protect credibility more. AI can create a case study in a minute. It cannot decide whether the story deserves to exist.
That is where human judgment must get stronger.
Earned media will reward the teams that still know how news works
AI search may change how people discover information, but it does not erase the old rules of credibility.
Reporters still need real stories. Editors still care about relevance. Readers still recognize fluff. Sources still need something useful to say. A brand still earns attention by contributing facts, perspective, access, or proof that others do not have.
The teams that win in this era will not be the ones that simply generate more content. They will be the ones that understand what public evidence looks like and how it gets created.
That means the training question is now a strategic question.
Who trains the next generation of PR pros in the AI era?
Agencies do. In-house teams do. Senior practitioners do. Clients do, when they give young people access to the substance behind the story. Reporters do too, indirectly, every time they ignore a bad pitch or respond to a good one.
But the industry cannot rely on slow tools to teach fast learners anymore.
We have to be intentional. We have to let junior people use AI, then ask them to prove their thinking. We have to give them real exposure to reporters, real edits, real stakes, and real explanations. We have to teach them that the work is not the document, the list, the pitch, or the coverage report.
The work is learning how attention is earned.
The fax machine is gone. The clipping book is gone. The Thursday trade-mail verdict is mostly gone too.
The need for judgment is not gone. It is more valuable than ever.
For more info: BestPRguy@gmail.com 310 920-2424 www.PublicityandMarketing.com
Comments