The prospect of never having to read another news story may become a reality with the use of artificial intelligence capable of analyzing web content and delivering summaries when requested. This scenario presents significant challenges for media leaders as companies like Google and others explore generative AI, a technology that generates fresh content based on past data.
In an image taken on June 9, 2022, at the Arena Berlin in Berlin, Germany, a woman walks by the emblem of the web search engine company Google during the digital society festival ‘re: public’.
In May, Google initiated the release of a novel search feature driven by generative AI. This move came after concerns were raised about the company’s future role in delivering information to consumers, given the growing popularity of OpenAI’s query-answering chatbot, ChatGPT.
Google’s product, known as Search Generative Experience (SGE), employs AI to produce summaries when specific search queries are made, based on Google’s assessment of whether a summary would be beneficial. These summaries are prominently displayed at the top of Google’s search homepage, with accompanying links to explore further, as outlined by Google’s description of SGE.
Publishers looking to safeguard their content from being utilized by Google’s AI for generating these summaries must employ the same tool that prevents their content from appearing in Google’s search results. This approach effectively makes their content nearly invisible on the internet.
For example, when conducting a search query such as “Who is Jon Fosse,” the recent recipient of the Nobel Prize in Literature, Google’s search results yield three paragraphs of information about the author and his literary contributions. Dropdown buttons offer convenient access to Fosse-related content on various platforms, including Wikipedia, NPR, The New York Times, and other websites. Furthermore, supplementary links are situated to the right of this summarized information.
Google clarifies that these AI-generated summaries are constructed by aggregating content from multiple web pages, with the accompanying links intended as a starting point for users seeking more in-depth information. Google characterizes its Search Generative Experience (SGE) as an optional experiment for users, aimed at enhancing and refining the product while integrating feedback from news publishers and other stakeholders.
For publishers, the introduction of the new search tool represents another challenge in their decades-long relationship with Google. In this relationship, both sides have grappled with competition for online advertising while simultaneously relying on Google for web traffic.
The evolving product, now accessible in the United States, India, and Japan, has raised concerns among publishers who are grappling with their position in a world where AI could play a dominant role in how users access and pay for information. Four major publishers, who requested anonymity to avoid complicating ongoing negotiations with Google, expressed concerns regarding web traffic, the proper accreditation of information in SGE summaries, the accuracy of these summaries, and compensation for the content that Google and other AI companies use to train their AI tools.
Google responded, stating, “As we bring generative AI into Search, we’re continuing to prioritize approaches that send valuable traffic to a wide range of creators, including news publishers, to support a healthy, open web.” Regarding compensation, Google is working to better understand the generative AI business model and is seeking input from publishers and other stakeholders.
In late September, Google introduced a tool known as Google-Extended, which allows publishers to prevent their content from being used to train Google’s AI models. While this is viewed as a “good faith gesture,” it raises questions about whether payments will follow and to what extent there is openness to establishing a fair value exchange.
Publishers are also concerned that SGE may reduce their organic traffic, as users may rely on the summarized information provided by SGE rather than clicking on links, leading to potential declines in click-through rates. Publishers have spent years optimizing their websites for traditional Google search but lack the information needed to do the same for the new SGE summaries.
Google asserts that publishers do not need to make any changes to appear in SGE summaries, as they have already allowed Google to “crawl” their content for search purposes. However, publishers contend that Google hasn’t been transparent about how they can block content from being crawled for SGE, and they express apprehensions about Google using their content to create summaries that may deter users from visiting their sites.
Overall, publishers’ concerns revolve around the perception that Google is using their content to create summaries that users may read instead of clicking on their links, all without clear guidance on how to block such content from being used in SGE. This situation raises questions about the potential impact on web traffic and publishers’ efforts to adapt to this new AI-driven landscape.
These concerns come at a time when publishers have long been navigating the complexities of their relationship with Google, which has implications for online advertising and search traffic.
Exclusive data from AI content detector Originality.ai reveals that websites are more inclined to block their content from being used for AI if it doesn’t affect their search results. Since the release of Google-Extended on September 28, 6% of top websites have blocked it, while 27.4% are blocking ChatGPT’s bot, including major publications like The New York Times and The Washington Post.
Overall, the introduction of SGE by Google has raised numerous questions and concerns within the publishing industry, adding complexity to an already intricate relationship between publishers and the tech giant.
