Generative search is only the beginning. What comes next as AI, monetization, regulation and shifting consumer behavior continue to reshape how brands earn visibility, trust and influence?
At the end of 2025, Muck Rack hosted the first-ever Generative Pulse Summit. More than 100 communicators came together for a day of education and networking about all things Generative Engine Optimization (GEO).
Communications leaders from ProRata.AI, Bloomberg and The Weber Shandwick Collective came together to look ahead to what the future may look like as GEO continues to evolve.
The experts included:
- Annelies Jansen, chief business officer, ProRata.AI
- Chaim Haas, head of innovation communication, Bloomberg
- Dotan Limon, chief product officer, The Weber Shandwick Collective
The truth of the matter is, we’re in the early days of GEO. While we’ve made incredible strides over the last 12 months, this technology is still extremely new, with lots of room to change, shift and grow.
“No one can be held responsible for the guesses that they make in terms of how the future goes here,” said Peter Mertens, Muck Rack’s chief of staff and moderator for this panel.
🔮 Where will Generative Engine Optimization land within an organization?
A major theme has emerged in Muck Rack’s AI research—and throughout all of the panels during the Generative Pulse Summit: GEO is ripe for PR to own, the way our marketing counterparts took hold of SEO years ago.
But where will GEO ultimately land within an organization?
“I would advise everybody to be open-minded and to really try to understand what GEO means for your organization today,” said Annelies Jansen, chief business officer for ProRata.AI. “And in some organizations, it can be owned by one department. In others, it needs to go much wider.”
Our panelists agreed it will likely need to be an organization-wide effort, with many departments collaborating.
“In many ways, earned media is driving GEO,” said Chaim Haas, head of innovation communication at Bloomberg. “So given that, comms certainly should have a stake at the table.”
However, think about who owns other important areas of your organization that impact GEO, like website, SEO, etc.
“Maybe it's a cross-functional task force that's responsible for all of these things, at least at this stage, because you have to see, how do you muster all the different elements so that your brand shows up the way you want [it to]? Because there are so many different pieces of it,” Chaim said.
Dotan Limon, chief product officer for The Weber Shandwick Collective said that while right now marketing and communications is poised to own GEO, he does see it evolving as AI becomes more advanced.
“When we think about where GEO is going, it's not just people reading or getting answers. It's ‘I want an AI agent to book a trip for me,’” he shared as an example.
For this reason, Dotan believes GEO will go way beyond marketing and communications and into end product integration and beyond.
🔮 What will monetization look like inside LLM responses?
For Chaim, monetization is not new, so he’s not opposed to experimenting with paid content inside LLMs.
“Many of the brands we know as earned media also do paid media or sponsored content, and for communications professionals, that's not a new world for us,” he said. “It's an interesting tactic to try to see what that does in terms of advancing your brand story with the LLMs, and whether or not that is a long-term strategy.
The bigger challenge, Chaim says, is understanding the technology behind the models and being aware of changes as new models emerge, because strategies will need to shift and change.
Annelies cautions communicators to not only pay attention to the five big AI companies, but to the “torso and tail AI solutions”—decentralized AI solutions used by millions where your brand has the opportunity to show up.
“Advertising, in our view, will power a large part of it, because the advertiser follows the user, and 1 billion users need to see ads,” said Annelies.
“I do think there will be two sides,” said Dotan. “There will be the brand side that will pay for something. The second is how publications will monetize, so there needs to be more of a value exchange here to make this ecosystem still work.”
Transparency in sharing when AI responses are sponsored and when they are not will become increasingly important for users to maintain trust with the LLMs.
🔮 How can communicators balance the efficiency we gain from AI with potential reputational risk of overautomating?
There’s definitely a balance between the work AI can take on and the work that must be done by a human, especially for those working in public relations and communications.
"The way I think about it in my day-to-day practice is: what are the tasks that usually take me a lot of time, and how can I speed those things up using these tools?” Chaim shared.
He shared the example of using it to speed up writing press releases, which can often be a time-consuming task.
“When I first started in this industry, I was told ‘spend an hour writing every press release,’” he explained. “Well, if I have to write 30 press releases, that's 30 hours and it adds up. But if I can take a recording of the event I'm supposed to be writing a press release about and throw it into multiple GPTs, I can get three different press releases, and then pick and choose the best parts and put them all together into the version I like and want to put out into the world. That all takes 20 minutes and becomes way more efficient.”
Chaim said in this case, he’s still the human in the loop—the decision maker at the end of the day who decides how the brand is represented to the outside world.
Dotan said one of the risks with using AI is that everything may end up sounding the same, or very average. What’s to differentiate your brand from someone else’s?
The key here, Dotan shared, is ensuring you’re giving the LLM you’re working with the tools it needs to help you create great content—you can upload brand voice documents, past press releases, speeches, transcripts, etc.—to help the content feel more distinctly you.
Annelies agreed with Dotan’s approach and compared it to hiring a summer intern and giving them an assignment without sharing much context. The end result is not going to be very good, so what you—the human—inputs into an LLM matters.
But, be cautious how much you share with an LLM. Never upload sensitive data or private information you wouldn’t want the public to see.
“We're privy as PR people, to a lot of information that's not yet public,” Chaim said. “Think about that as you're using these assistants.”
🔮 What’s the next big AI shift we’ll see that comms pros may not be thinking about yet?
🎤 Annelies: “I think it's paid. Brands want it. Brands want to spend money in AI, because that's where the users are. How do we, collectively, as an industry, make that paid experience a safe one? I would definitely start to think about that as the industry, because now it's time to set the guardrails for our brands and for our clients.”
🎤 Chaim: “Agentic AI. We're just starting to see that on a consumer basis where you can start to ask the agents to go out and do tasks for you: order from the grocery store, book my travel, etc. That's your customer experience. And how do you control that if these agents are doing all that stuff on your website that you've invested so many millions of dollars on? What happens when the agent runs into a problem? Those are the things brands need to start to think about. And even agentic AI is still in its infancy. The standards are still being hashed out. There's a lot of work to be done there.”
🎤 Dotan: “Technology has become almost a dirty word in communications. We as communicators need to become more immersed in technology. We need to find ways to create a knowledge management system about our brands from websites. And it's very technology heavy, but we have to really think about how we manage that knowledge graph about our brand internally and externally, and how we identify those gaps in knowledge.”
For more, check out a full recap of the Generative Pulse Summit, including audio from each session—and be sure to download the latest What Is AI Reading? report.
