Photo by Steve Johnson on Unsplash
 

I’m in Cannes this week for the Cannes Lions International Festival of Creativity. It’s a week-long celebration of excellence in branded communications and advertising. So, as you can imagine, the event is filled with people who create, buy, and sell advertising. As you can also imagine, every conversation ultimately gets to the subject of AI. Among the pressing concerns, one of the biggest is the threat that web-connected large language models (LLMs) pose to web publishers. Let’s review.

They are variously called “Web-Connected LLMs,” “Web-Enabled LLMs,” or even “Internet-Augmented LLMs.” Whatever you like to call them, they are AI models with the ability to browse the web. This is a fantastic feature for users who need summaries of timely information. For advertiser-supported publishers, it’s the beginning of the end of days – LLMs don’t click on ads.

Impacts on Advertising Revenue

Bot Traffic Up, Human Traffic Down: Web-Connected LLMs provide users with direct answers, reducing (and in many cases eliminating) the desire/need for users to visit original sources. This decrease in web traffic directly impacts ad impressions and click-through rates.

Ad Evasion: Generative AI models do not interact with web ads, bypassing the primary monetization mechanism for many content sites. This means fewer opportunities for ad views and clicks, diminishing revenue streams.

Content Scraping Concerns: LLMs tend to surface publisher-scraped content without proper attribution or compensation. Once content has been scraped, the publisher has lost control of its intellectual property – so future monetization gets much harder (to say the least).

Some Strategies for Adaptation

Subscription Models: Publishers can attempt to shift to subscription-based models, offering premium content and services behind paywalls. High-quality, exclusive content can drive subscriptions and reduce dependency on ad revenue. But, as we all know, a dual revenue model (ad-supported plus subscription) is the industry standard, and nowadays it’s a requirement to run a healthy content business.

Direct Monetization: Developing new direct monetization strategies, such as microtransactions for articles or implementing more robust membership programs, can provide alternative revenue streams. Remember when Jeff Zucker was CEO of NBCU? It was circa 2008 when he said, “We’re exchanging analog dollars for digital dimes.” Back then, he was talking about the transition from network to networked television (I wrote a book about it in 2006). His quote was prescient and absolutely accurate. For the current conversation, we might update it to: “We’re exchanging digital dimes for AI pennies.”

Licensing Content: Publishers have already started to license their content to AI companies, ensuring they receive compensation for the use of their material. But this seems like a very short term solution. Once ingested, it can’t be uningested – no additional access to the publisher’s data is required.

Regulation

Publishers and content creators are already engaging with policymakers to establish regulations that protect content creators and ensure fair compensation from AI companies. This could involve copyright reforms or new legislation tailored to the Generative Era. There are also a lot of pending lawsuits that may help shape the future of content scraping. But, so far, none of these issues is close to being settled.

What’s Next?

The rise of web-connected LLMs is rapidly undermining traditional web publishing. It’s clear industry professionals are deeply concerned. LLMs reduce human web traffic, evade ads, and scrape content without proper attribution, all of which erode publisher revenue. Strategies like subscription models, direct monetization, and content licensing offer some revenue mitigation, but they are insufficient. Ongoing regulatory efforts and legal battles may eventually provide some relief, but these measures are far from resolved.

What’s next? My guess is that the big publishers will get bigger, the small will get smaller, and the middle will disappear. We’ll see the biggest publishers strike multimillion dollar deals, recede behind paywalls, and continue to create unique, high-quality content. Everywhere else, consumers will be inundated with synthetic content (which will come with varying levels of value). Said differently, I don’t think this is going to end well for the medium-sized web publishers, and it may not be much better even for the biggest. We are already being inundated with synthetic content – the amount of this content will continue to increase exponentially. At some point, even the biggest publishers will be overwhelmed by synthetic competition. Clearly, the biggest publishers are also in the best position to take advantage of the technology. Will they?

It’s impossible to predict the future, but it’s very easy to predict the continuation of this series of tech trends that pose extraordinary headwinds for the financial structure of publishing.

More By This Author:

Apple WWDC 24 AI Summary
Apple Intelligence: Apple’s AI Coming Soon
OpenAI Introduces ChatGPT Edu