Moderating a panel of CEOs from Gannett, McClatchy, and The Atlanta Journal-Constitution at the INMA World Congress of News Media last week, I asked them their biggest regret from the last year or two.

I rarely like it when all panelists give the same response, but what struck me is that maybe every CEO would say this: I wish I’d acted on AI sooner. 

AI came up in nearly every discussion at World Congress. It’s no longer theoretical. It’s reshaping everything.

I break it down into four buckets:

1. Business efficiency was the most mentioned. There are mass efficiencies of scale as AI allows for repetitive tasks to be automated. This will show up all across the business from automated print layouts, media planning, marketing, summarising, headline writing, etc.

2. Product innovation as AI develops. This is the growth area. The obvious ones to note are personalisation of content and format, an area I suspect we’ll see more of over the next year. One example that jumped out to me was The New York Times’ brand match where it uses AI to help marketers quickly find the best content for their needs. Expect more inventive use cases to show up as initial experiments mature.

3. Changes to consumer behaviour is still an unknown. It’s now common wisdom that search is (in some markets) or will (in others) decrease and in some cases fairly rapidly. I’ve been told that publishers should prepare for near-zero search traffic within two years. Will answer engines replace search? Which new interfaces will emerge? And how wary will consumers become about AI-driven truth claims? We simply don’t know yet.

4. AI partnerships and licensing: During my 2.5-hour Product & Tech seminar with OpenAI, Perplexity, Microsoft, Tollbit, and Human Native, it became clear that deals vary wildly — and none are built to last. AI firms have lean teams. We heard from media execs that minimum guarantees are important. Perplexity still faces pushback over no/low ad revenues versus high content-placement risk. Meanwhile, many media outlets remain oblivious to where bots are scraping their work (blocking them is the first step). 

OpenAI kind of announced that ChatGPT intends to integrate media subscriptions and let users set source preferences. I say “kind of” because it’s in development and not in production with any idea of release. I found this promising as it seems there are serious conversations about how subscription content might be presented in AI answers, including layout and branding. 

So what should media leaders do now? I asked Madhav Chinappa and Josh Stone, both tech veterans now at AI startups, for practical advice:

1. Get your house in order internally. Data architecture, metadata tagging, content workflows — nail the fundamentals.

2. Audit your content’s footprint. Know exactly how, where, and by whom your work is being scraped. Tollbit was on the panel. ScalePost and Cloudflare also have service that allow for this.

3. Create true scarcity. If your journalism is everywhere, why should anyone pay for it? Carefully manage distribution to preserve value.

AI isn’t a distant horizon. It’s today’s frontier. The device landscape, consumer habits, platform relevance — even the open Web’s future — are all up in the air. By shoring up our foundations, we can move from reactive regret to proactive advantage. 

If you’d like to subscribe to my bi-weekly newsletter, INMA members can do so here.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .