If artificial intelligence is able to process all the information on the internet and provide a summary whenever needed, you might never have to read another news story in your life.
For media magnates, that’s the stuff of nightmares as Google and other companies experiment with generative AI—a technology that generates new content by using historical data.
After industry watchers questioned Google’s future prominence in providing consumers with information in the wake of the rise of OpenAI’s query-answering chatbot, ChatGPT, the tech giant started implementing a new type of search powered by generative AI in May.
The product is named Search Generative Experience (SGE), and it leverages artificial intelligence (AI) to generate summaries in response to certain search queries based on whether Google’s system thinks the format would be beneficial. According to Google’s overview of SGE, those summaries are displayed at the top of the search engine’s homepage along with links to “dig deeper.”
To stop their content from being utilized by Google’s AI to create those summaries, publishers have to use the same tool that would keep their content out of Google search results, making it almost completely invisible online.
When you search for “Who is Jon Fosse,” the recent winner of the Nobel Prize in Literature, for example, you get three paragraphs about the author and his writing. Additional links appear to the right of the summary. Drop-down buttons link to Fosse content on Wikipedia, NPR, The New York Times, and other websites.
According to Google, the links in the AI-generated summaries are meant to serve as a starting point for further information. The summaries are reportedly compiled from a variety of web pages. To help it evolve and improve the product, it describes SGE as an opt-in experiment for users, incorporating feedback from news publishers and other sources.
Publishers have struggled to compete with Google for online advertising and have relied on the tech giant for search traffic for decades. The new search tool is the latest red flag in this relationship.
According to four prominent publishers, the still-evolving product has raised concerns among publishers as they try to figure out their place in a world where AI could dominate how users find and pay for information. The product is currently available in the United States, India, and Japan.
According to those publishers, the concerns are about web traffic, whether or not publishers will be given credit for the information that appears in the SGE summaries, and how accurate those summaries are. The primary issue with AI is that publishers want payment for the content that Google and other AI companies use to train their tools.
In a statement, a Google representative stated that in order to maintain a strong, open web, we’re continuing to give priority to strategies that drive valuable traffic to a variety of creators, including news publishers, as we integrate generative AI into Search.
Regarding payment, Google claims that in order to improve its comprehension of the generative AI applications’ business model, it is seeking feedback from publishers and other stakeholders.
Google debuted Google-Extended, a new tool that allows publishers to prevent their content from being used by Google to train its artificial intelligence models, in late September.
According to Danielle Coffey, president and chief executive of the News Media Alliance, an industry trade group that has been lobbying Congress on these issues, offering publishers the choice to opt out of being crawled for AI is a good faith gesture. It is unclear to what degree there is openness to having a healthier value exchange and whether payments will follow.
Publishers cannot disable the summaries or the links that accompany them from being crawled for SGE using the new tool without having their content removed from standard Google search results.
In order to attract advertisers, publishers need clicks, and being visible in Google searches is essential to their operations. According to an executive at one of the publishers, the SGE design has moved the links that show up in traditional search further down the page, potentially reducing traffic to those links by as much as 40%.
Even more concerning is the potential for web browsers to choose not to click on any of the links if the SGE passage satisfies their informational needs—for instance, by providing information on the ideal time of year to visit Paris without requiring them to visit the website of a travel magazine.
According to Forrester Research Senior Analyst Nikhil Lai, SGE will undoubtedly reduce publishers’ organic traffic, so they will need to consider alternative metrics to gauge the content’s worth instead of relying solely on click-through rate. However, he thinks that despite their links appearing in SGE, publishers’ reputations will continue to be strong.
According to Google, SGE was created to draw attention to web content. According to a company spokesperson, any estimates regarding the precise traffic impacts are tentative and not representative, since what you see in SGE today might not resemble what eventually launches more widely in Search.
Publishers and other industries have spent decades optimizing their websites to appear prominently in traditional Google search results; however, these publishers claim they lack the necessary data to do the same for the new SGE summaries.
For us, the new AI section is a mystery,” a publisher’s executive declared. We don’t know the algorithm underlying it or how to confirm that we are a part of it.
According to Google, publishers can continue doing what they are currently doing to appear in search results.
For a considerable amount of time, publishers have permitted Google to automatically scan and index their content through a software programme known as a “crawl,” or bot, in order to display it in search results. Google uses a process called “crawling” to index the web and display content in searches.
The main concern raised by publishers about SGE is this: they claim that Google is freely crawling their content in order to generate summaries that users can read instead of clicking on their links, and they are unsure of Google’s policy regarding blocking content from being crawled for SGE.
According to a publisher, “Google’s latest search tool poses a greater risk to our company and its operations than an unauthorized crawler.”
Google remained silent regarding that evaluation.
Exclusive data from AI content detector Originality.ai shows that when given the choice, websites are refusing to allow their content to be used for AI if doing so has no effect on search. 27.4% of the most popular websites, including The New York Times and Washington Post, have blocked ChatGPT’s bot since its release on August 7. In contrast, since Google-Extended’s release on September 28, 6% of users have blocked it.
