John Croll - author I’ve been in this industry a long time. I’ve watched it move from press clippings to PDFs to dashboards. Every cycle promised something new and mostly delivered more of the same: a daily report sitting in someone’s inbox at 6am that was automated from the platform or junior analyst took two hours to compile. That era is evolving. When I talk to founders and chief comms officers, the conversation has changed completely in the last three or four months. Twelve months ago they were asking us for better summaries, faster turnaround, prettier reports. Today they’re asking how our AI tools work, what data they can interrogate, and whether we can help them get ahead of a story when it breaks. They’re not buying media monitoring anymore. They’re buying data intelligence. That’s the shift. Daily reports are an audit. By the time you read them, the moment has passed. Data intelligence is real-time, queryable, and actionable. You can do something about what you’re seeing. That’s the difference between counting clippings and shaping coverage. The productivity dividend (and the value that comes with it) Here’s where a lot of agencies get it wrong. The industry is under genuine pressure right now. AI in media monitoring has compressed the work that used to take humans days into a few minutes. Clients see it. They’re asking for productivity dividends, reasonably so. ADVERTISEMENT But the assumption baked into that conversation, that automating the analyst’s work makes the service less valuable, is backwards. If a $20,000 quarterly insights agreement now produces the same output in a fraction of the time, the obvious response is to cut the price.  The smarter response is to ask what else becomes possible when the work happens in real time instead of weeks later. When you’re not waiting for a quarterly deck, you can intervene mid-campaign. You can spot adverse coverage trending before your CEO sees it on LinkedIn. You can brief a journalist Tuesday morning on data that didn’t exist Monday night. There’s a productivity dividend, yes. But there’s also a value appreciation, because information that arrives at the moment you can act on it is worth more than information that arrives when the action window has already closed. Where comms leaders are using AI well The teams getting this right aren’t running ChatGPT for press releases off the side of their desks. They’re using generative AI for communications strategy work, the parts that used to be reserved for the most senior person in the room. Campaign frameworks. Issues management playbooks. Target audience impact. Journalist mapping. The first draft is now genuinely useful, and the human time goes into pressure-testing it against what they know about the client and the market. The other thing that’s changed: prompts on data platforms. The 6am daily news brief, the one your senior analyst hates writing and your executives skim in the lift, can now be generated, customised by the audience, and delivered before anyone’s awake. Outsourcing it is expensive. Doing it yourself is exhausting. Letting AI tools for PR work the data set is neither. There’s also a quieter shift worth flagging. In B2B, around 82% of citations in AI-generated answers are coming from earned media. If your content is well-structured and your coverage holds up, large language models will index you and cite you. PR has always been about credibility at scale. AI just reset the channel. Where they are tripping up using AI The same place every industry trips up. One-prompt strategies, media releases pushed live in a single pass, crisis statements that read like they’ve been generated by a chatbot — because they have. No depth, no second draft, no editorial judgement applied. The pattern is consistent. The comms leaders who use AI well treat the model as a junior on the team: fast, eager, occasionally wrong, and absolutely requiring a second set of eyes. The ones who get burned treat it as the answer. What the next twelve months looks like Two things I’m watching. First, AI capability is now a procurement filter. Top-down mandates have pushed AI into the buying criteria, and vendors who can’t show one are being cut from shortlists. Quietly, but firmly. Second, the trust threshold has risen. Comms teams want AI working on their own data, not summarising the open internet. The platforms that get this distinction right will own the next decade.  Related Content Opinion Going back on your word is riskier in an AI world Events CommsCon catchup: Humans still required — communication in an age of misinformation SUBSCRIBE Sign up to our free daily update to get the latest in media and marketing. Topics ai ChatGPT John Croll media monitoring Truescope Have your say Login with:Continue with LinkedIn Or comment anonymously Your comment will be marked as unverified Name Comment Δ Latest What I'm hearing from comms leaders about AI Things looking up for Ooh Media on the Unmade Index Initiative CEO Jo McAlister jumps to News Corp ARN appoints Lucy Formosa Morgan in new sales director role The 2.5 second survival guide for advertisers