Google’s artificial intelligence could kill internet search and publishers’ businesses

Google is holding its annual Google I/O developer conference today, and there is growing concern among webmasters about the expanded capabilities of Google’s new artificial intelligence search tool, The Washington Post reports. This feature, called Search Generative Experience (SGE), directly responds to user queries with comprehensive, multi-paragraph answers, pushing links to external websites further down the page and reducing their visibility.

This development could significantly disrupt the web system. Millions of authors and publishers who depend on Google traffic may be at risk of losing their business. Some experts argue that the inclusion of AI will further strengthen Google’s already significant control over the Internet, leading to a scenario where information is provided mainly by a few large corporations.

Google calls these AI answers “reviews,” but they often paraphrase content directly from other websites. For example, a search for tips on how to fix a leaking toilet yielded an AI-generated answer that included the advice to tighten the bolts of the cistern, taken directly from an article on The Spruce, a home repair website owned by Dotcom Meredith. However, the link to the AI source was minimal, and the original content was partially hidden and required additional clicks to view.

According to projections by Gartner, web traffic from search engines will decrease by 25% by 2026. Ross Hudgens, CEO of SEO consulting company Siege Media, predicts a 10-20% drop in traffic for many publishers, and some will face an even bigger drop. “Some businesses will be killed,” he said.

Raptive, a company that provides digital media services to about 5,000 websites, estimates that Google’s search changes could result in losses of about $2 billion for content creators, with some sites losing up to two-thirds of their traffic. Michael Sanchez, co-founder and CEO of Raptive, warns that these changes could cause “enormous damage” to the structure of the Internet, jeopardizing the survival of an open and diverse network.

Despite these concerns, Google maintains that it intends to prioritize website traffic. CEO Sundar Pichai emphasized during the financial statement call that the company is acting cautiously to ensure the health of the web. However, a Google spokesperson declined to comment further.

Selena Deckelmann, Director of Product and Technology at Wikimedia, reflects on the broader implications, noting: “We’ve seen an incredible growth of the Internet… Now we’re at a point where I think profit is pushing people in a direction that I’m not sure makes sense.”

Jake Boley, a strength and conditioning coach in Austin, noticed a 96% drop in traffic to his workout shoe review website last year. Google still refers to his content in AI-generated answers, but as a result, fewer users are visiting his site. “My content is good enough to scrape and summarize,” he said. “But it’s not good enough to show in your normal search results, which is how I make money and stay afloat.”

Google started experimenting with generative AI in search after the release of ChatGPT by OpenAI. Microsoft, Google’s main competitor, integrated an AI chatbot into its search engine, prompting Google to follow suit. Subsequently, Google implemented AI technology in various products, including Google Docs and YouTube video editing tools.

Search remains Google’s most important product, accounting for 57% of the company’s $80 billion in revenue in the first quarter of this year. Although AI answers are mostly not shown on popular queries that may contain ads, this feature may force websites to buy ads to maintain their search rankings.

At the same time, the integration of AI into Google search is not without problems. Answers generated by artificial intelligence can be incorrect and sometimes fictitious. Training artificial intelligence on a large array of content, often without permission or payment to the original authors, has already caused legal problems. OpenAI and Microsoft have faced lawsuits over alleged misuse of copyrighted works.

Frank Pine, executive editor of MediaNews Group, equates the use of copyrighted content by artificial intelligence with plagiarism: “If journalists did this to each other, we would call it plagiarism.” He warns of catastrophic consequences for the journalism industry if tech companies continue this practice without proper compensation.

Despite these challenges, the momentum for AI-assisted search and centralization of web information is not fading. OpenAI has already signed agreements with web publishers to place their content in its chatbot, which indicates a potential change in the way information is accessed and distributed online.