How AI could supercharge ‘go direct' PR, and what the media can do about it

Fast Company

23-05-2025

In the past several years, the trend of 'going direct' in public relations has gotten trendy. Broadly, the idea is that certain companies—mainly tech startups—stand a better chance of advancing their own narratives by sidestepping traditional PR and media altogether. Instead, the company founder, fellow executives, and partners would post content to the internet and social media to directly communicate with their customers.
There's naturally been a lot of consternation in the media and PR industries about how effective this kind of approach is, the real value of traditional PR, and whether a company can really chart their own path without some kind of third-party validation. It's not my intent to wade into that debate (though if you'd like a deep-dive exploration, I hosted a panel on the topic at the Consensus conference). It is, however, an undeniable trend that's caught fire the last few years.
Now AI is poised to throw gas on that flame.
The next evolution of going direct
I was struck by this after reflecting on my conversation with Scrunch AI CEO Chris Andrew on The Media Copilot podcast. Scrunch specializes in placement in AI search. Its customers are mostly brands who want to ensure their content is crawled, analyzed, and summarized when someone asks a chatbot about the brand or its area of focus. The idea is conceptually similar to SEO (search engine optimization), though the industry hasn't yet settled on a name for it (AIEO, LLMO, and GAIO are all contenders).
As Google has just aptly demonstrated in its push this week to elevate AI Mode as a standard feature in search, the purpose of an AI search is to give answers, not links. That's a huge problem if your product is information, which is exactly why much of the media industry is locked in a legal battle with the AI industry over copyright. But if you're a company just trying to sell something, an AI summary that informs a user about your brand is a win whether they click through to your site or not. If they do, there's information to suggest they'll be much more inclined to engage further and even transact. And if they don't, you've effectively hit them with an ad by having the brand mentioned in the summary.
On the media side of things, the click-killing aspect to AI search has many outlets throwing up defenses on their content against crawlers. They're configuring their robots.txt file to say 'no' to bots, putting up other digital defenses, and denying access to their content unless AI companies pay up—either through licensing agreements or pay-as-you-go frameworks.
A recent story in A Media Operator, which covers the business side of the media industry, showed that many media companies have begun to wake up to the rapidly growing presence of AI crawlers. An executive from Cloudflare, an internet infrastructure company, said over 800,000 websites have activated Cloudflare's most aggressive protection setting.
There's an obvious disconnect between the incentives of the media versus brands in AI search, and that creates an opportunity for an AI upgrade of the go-direct strategy. AI search engines still need to provide answers to queries, and if credible journalism about those topics is blocked, something has to fill the void. Brands that give unfettered access to their content to crawlers (because why wouldn't you?) will have an advantage.
This goes double if the company can execute on a multichannel content strategy that gets their brand cited across multiple sites or domains. One important difference between AI search and SEO, pointed out by Andrew in the podcast, is that citations count more to AI crawlers than links. That means if a brand can seed the web with consistent facts and brand citations across multiple sites, it will help ensure AI search engines 'learn' from their preferred narrative.
You can imagine a scenario where a major company, with enough resources, could theoretically pull off a version of what Russia has done with respect to advancing their preferred narrative on the war in Ukraine, thoroughly examined by a NewsGuard investigation. Except in Russia's case, it was done mainly via sketchy-looking sites clearly created to 'spam' AI crawlers with propaganda. A company could do this out in the open, with a content strategy that amplifies their storytelling across blogs, podcasts, social media, and more, published across multiple domains. Humans would easily be able to tell it's all marketing, but AI engines just see it as more data—data that can have a large amount of influence in what appears in summaries.
How the media can chart a smarter course
There's still hope to steer away from a future where corporate propaganda is dominant. It starts with media sites adopting a sophisticated approach to blocking, something I outlined in my newsletter last week. Blanket bans are understandable—publishers still feel burned by Big Tech's platform dynamics of the past—but shutting off access entirely is a short-term defense with long-term costs. A more strategic approach would involve selectively exposing certain types of content: meta descriptions, older articles, multimedia, and more. This allows media companies to remain visible in AI search while still protecting core value.
But beyond technical solutions, the real hope is in what consumers of information actually want. Review sites like PCMag and The Wirecutter didn't become popular because they were algorithmically boosted. They emerged because people didn't like getting fed the company line. Similarly, if AI-generated answers start to feel like corporate brochureware, consumers will notice. Credible, independent journalism isn't just good ethics; it's a market advantage—if it's accessible.
In the end, AI engines that optimize for this balance will win out, too. It's right there in ChatGPT's model spec: the chatbot is designed to 'seek the truth together' with the user. It can't do that without including independent perspectives and weighing them appropriately against a barrage of go-direct content.