A touch of tech in comms — the future of communication powered by AI

AI is already reshaping how brands inform and are cited: from SEO to GEO, from static content to modular structures, from chatbots to agents. In this edition of the 5 Pontos newsletter, we begin the discussion on the practical steps to make data machine-readable without losing the human voice and to govern reputation within artificial intelligence models. Enjoy reading!

1. GEO: From SEO to being cited by AI

With a projected 25% drop in traditional search volume by 2026, the competition is no longer just about rankings and clicks; it also runs through the synthesis of generative models.
The goal isn’t merely to “show up”: it’s to be cited as a trusted source. That calls for a gradual shift from purely narrative press releases to a more citable core: facts, proprietary data, historical series, transparent methodology, and machine readable metadata.
Here’s a practical point: signals of freshness, consistency, and sourcing increase the odds your brand is treated as a reference in answers and summaries. In GEO, winners tend to be consistent and citable—not necessarily the most popular. Put simply, brands must become providers of context for their own story. If they don’t leave structured, verifiable traces, they’ll be underrepresented—and open the door to incomplete syntheses.

2. Modular content: One story, many format

Journalism’s new era already works with knowledge blocks: the same story becomes a short video, a newsletter summary, and data packages for dashboards. This modularity isn’t only about distribution; it’s about being ingested by newsroom engines and by AI models themselves. In a landscape where 40% say they avoid news due to fatigue and overload, audiences want smaller, more direct, reusable doses. And in many markets—especially among younger audiences—social and video have become central to discovery, including via personalities. The solution is to start from a clear factual core and produce derivative versions that are consistent with one another, each with its own URL and metadata. Without semantic structure, your content becomes noise; with it, it becomes a recurring reference.

3. The use of agents in communication 

AI has moved from “writing emails” to coordinating processes. The next step is Agentic PR: agents that plan, research, monitor, and execute tasks under human supervision. The payoff is speed with control: teams focus more on strategy and relationships, while AI sustains volume, routine, and cross channel consistency—with governance. Today, 37% already use it to optimize content, 36% to draft external pieces, 35% for brainstorming, and 32% to customize pitches. These agents maintain live databases of journalists and KOLs, assemble pitches with data and human quotes, run real time clipping, and trigger crisis protocols based on predefined signals. On the back end, they consolidate reports with insights and actionable recommendations.

4. AI focus groups enable audience testing before launches

Instead of waiting weeks for traditional research, generative models now allow teams to simulate the perceptions of specific audiences — CFOs, journalists, regulators, or consumers — in a matter of hours. It becomes possible to anticipate the impact of a message, compare creative approaches, adjust tone and jargon, and run A/B testing at scale. The value lies in accelerating learning and reducing costly mistakes before major launches. The key, however, is to treat simulation as a hypothesis, not as truth. Prompts must be calibrated to the local context, analyses anchored in real market data, and findings validated with real samples whenever possible. Documenting assumptions and criteria helps mitigate bias. The result is continuous market intelligence, seamlessly integrated into the creative workflow.

5. Share of Model — the new reputation metric

Just as share of voice measured presence in the press, “share of model” is a way to think about how often—and in what context—AI models cite and recommend a brand. The contest isn’t for clicks; it’s for contextual authority: being the reference the system recognizes when it synthesizes. That demands a context architecture built on structure (machine-readable data), consistency across channels, and human authority signals—the AI seeks original sources to avoid “model collapse.” Ignoring this game brings three risks: reputational hallucinations, unmanaged associations, and strategic invisibility. The answer is continuous governance: publish data, keep facts live, link to verifiable sources, and correct what the AI gets wrong. Your reputation already exists in the model—who wrote it?

*This newsletter was produced with the assistance of artificial intelligence.