The AI Effect, Part 3: The AI search war is reshaping the web

This is the third in a three-part series on how AI is changing the ways consumers interact with the web and how advertisers can reach them across commerce, streaming and search. Read Part 1 and Part 2.  

Last week, the American Economic Liberties Project called on the U.S. Court of Appeals for the D.C. Circuit to, in part, strengthen remedies imposed on Google after it was ruled a search monopolist, and to revisit remedies that would “pry open” competition in search and protect competition in GenAI.  

The nonprofit’s amicus brief comes nearly two years after the Google search monopoly ruling in August 2024 and almost a year after Judge Amit P. Mehta ruled that Google had to share some search results with rivals but did not have to sell the Chrome browser.  

Mehta had focused extensively on how AI could alter search competition, writing in the remedies ruling that the “emergence of GenAI changed the course of this case.” 

Now, a structural shift of the search landscape is underway, with Google and a host of newer entrants like OpenAI locking horns. No matter who is “winning” the AI search war, the battle is already reshaping the open internet. It’s impacting how consumers interact with the web, putting pressure on publisher traffic and influencing how brands approach SEO and, now, GEO (generative engine optimization). 

“Consumers are shifting away from keyword searches and toward searches that are more conversational and specific,” said Kate Muhl, vice president analyst at Gartner. “This is especially true for younger consumers. This change was underway before the advent of AI chatbots, but conversational interfaces are driving that approach into the mainstream.” 

The power players right now 

The AELP wrote in its brief that the remedies decision had “erroneously interpreted a technological change as a competitive threat, instead of realizing that this inflection point heightened the urgency to restore competition.”  

Since that ruling, the struggle for dominance in AI search has ramped up, even as Google has regained momentum. In May, the company made what it called the search box’s “biggest upgrade in over 25 years,” fully integrating AI into the experience. Now, users can have full-on conversations, starting in the search box, with longer, more detailed prompts or questions. 

That same month, The New York Times declared that Google was “starting to win the AI race,” noting its changes to the search box, as well as a deal for Google’s Gemini to power Apple Intelligence features. 

But competitors have emerged, and AI search is taking many different forms, from conversational apps like ChatGPT to AI summaries in search platforms.  

OpenAI said in June that ChatGPT became the fastest app ever to reach 1 billion users. Anthropic’s Claude, more popular for business- and data-focused inquiries, has reportedly seen paying customers “skyrocket.” Others, like Perplexity and Microsoft’s Copilot, are vying for a slice of the pie. Even Reddit, which is a top source for AI citations, is powering its own search function with AI. 

Advertising comes for AI search and chats 

The next phase of the AI search race may be defined by advertising, as OpenAI tests ads in ChatGPT and Google experiments with new types of ad formats in Gemini.  

WPP Media projects global ad revenue from AI search could top $100 billion by 2030, more than traditional search and other channels. EMarketer projects AI ad spending to reach over $68 billion just in the U.S. over the same period.  

For now, however, the vast majority of that spend — 80% this year, according to EMarketer — will appear adjacent to AI content, such as ads next to AI search summaries, rather than directly inside AI chatbots or conversations. Those types of ads could still account for about 60% of AI spending by 2030. 

“Advertising on platforms like ChatGPT seems like a natural next step, because AI chat is its own kind of consumer media experience now,” Muhl said. “A search is a conversation. Answers come iteratively. And that’s by design, so that people will spend a lot of time with the chatbot. Brands will want to be where the engaged eyeballs are spending time.”

Still, Adam Ortman, CEO of Kinetic319, warned that brands shouldn’t treat AI search platforms and chatbots like traditional search because the distance between channel and user is collapsed: “AI feels so different from previous digital platforms because users increasingly treat it less like software and more like a personal adviser.”

“Advertising will almost certainly become part of the AI ecosystem,” he added. “The question is whether the industry introduces it thoughtfully enough to preserve the trust that makes conversational AI useful in the first place.”

Are brands optimizing or “sloptimizing”?

Success won’t just come from advertising inside AI environments. It will mean building a brand that is actually “worth finding,” Clare Farrugia, head of performance at Hatched, said at a recent event. That goes beyond advertising within AI environments and means reinforcing a brand’s discoverability.

“It’s going to become standard for every brand to track on a regular basis what AI is saying about them and then work backwards from there,” Gregory Galant, co-founder and CEO of Muck Rack, told The Current. “Why is it saying this and what can I do to influence it? That will be the biggest question on marketers’ minds for the next few years.”

But there’s a fine line between messaging for the AI era and what The Atlantic recently referred to as “sloptimization” — creating content primarily to appear in AI answers.

“Platforms that protect recommendation integrity could ultimately create more durable value than platforms that maximize monetization too aggressively,” Ortman said. “Brands that earn inclusion through expertise, authority and customer satisfaction may outperform brands that rely primarily on paid placement.”

Even given the AI hoopla, Muhl believes that it would be a mistake for brands to completely replace their non-AI approaches — because consumers haven’t.

“AI chatbots have been additive to the journey,” she said. “Consumers are adding AI chatbots to the list of sources they consult on the path from inspiration to purchase and post-purchase, not replacing steps or sources.”

Publishers feel the pressure

As advertisers race to monetize and strategize around AI search, the rest of the web is feeling the effects.

According to a Chartbeat report, page views from Google search dropped 34% from December 2024 to December 2025. Google rolled out AI Overviews to all U.S. users in 2024, and by mid-2025 it had expanded to more than 200 countries. Smaller publishers were hit the hardest, with page views declining 60%. 

And AI referrals aren’t offsetting the decline. While traffic from chatbots like ChatGPT continues to grow, it still accounted for less than 1% of total page views in the Chartbeat network as of the end of 2025.

“AI functions more like a research assistant than a news reader, directing people to how-to guides, explainers, and authoritative references,” the report added. “Publishers who balance timely coverage with evergreen, problem-solving content will be better positioned as this channel continues to grow.”

There are bright spots. According to a Muck Rack analysis of 25 million AI citations, journalism was the most-cited type of media at 27%, followed by corporate blogs and content at 24%.

Results varied significantly by query, platform and publisher. Galant described the landscape as the “Wild West.”

Winners are starting to emerge in AI citations

Different studies point to different winners.

According to a Tinuiti report, Reddit and YouTube account for the biggest share of AI citations of social platforms. According to Comscore, YouTube is the top cited domain in AI summaries for Google (which owns YouTube) and Bing across categories like entertainment, retail and sports.

For the Muck Rack analysis, Google’s Gemini cited Reddit (which has a data-sharing deal with Google) and YouTube the most, while ChatGPT cited Wikipedia and Axios most frequently.

However, Reddit CEO Steve Huffman recently told CNBC that “search is volatile” and Google’s AI Overviews have “yet to make a similar level of positive impact” on traffic as traditional search where links are more prominent.

LinkedIn was found to be the most-cited domain for professional queries across all AI search platforms in a recent report. Semrush similarly ranked Reddit and LinkedIn as the top two sources for LLMs like ChatGPT, Perplexity and Google AI Mode (Wikipedia was No. 3 and YouTube No. 4).

Galant believes that responses could improve, and increasingly favor quality journalism, as the AI search war continues (and as news outlets strike content licensing deals with AI companies).

“A bunch of AI companies are competing to win the consumer,” he said. “They all want to produce the best user experience. I don’t think anyone at those companies is sleeping at night thinking they’re the winner.”