The ongoing remedy phase of Google’s search antitrust trial has revealed how the internet giant used its stranglehold over internet search to strongarm publishers on the controversial topic of AI training. Internal documents show Google considering giving publishers an opt out on AI training before deciding to force publishers to feed their insatiable AI machine.
Bloomberg reports that as artificial intelligence continues to transform the search landscape, Google’s practices in using web publishers’ data to train its AI models have come under increasing scrutiny. A recent court document revealed during the company’s ongoing search antitrust trial sheds light on the tech giant’s decision-making process regarding publisher opt-outs for AI search features.
According to the internal document written by Chetna Bindra, a product management executive at Google Search, the company considered various options for giving web publishers control over how their data is used in AI search. These options ranged from allowing publishers to opt out of specific AI features, such as AI Overviews, to more comprehensive opt-outs that would prevent their data from being used to train AI models altogether.
However, Google ultimately decided against providing publishers with these choices, citing complications in training AI models and maintaining a seamless search experience. The document stated that Google had a “hard red line” and would require all publishers who wanted their content to appear in search results to also allow their data to be used for AI features. This decision was made without a public announcement, opting for a “silent update” instead.
The revelation has drawn criticism from industry experts and publishers who argue that Google’s approach limits their control over their own content and potentially harms their revenue streams. Paul Bannister, chief strategy officer at Raptive, a company representing online creators, described the document as “a little bit damning,” noting that it clearly showed Google’s awareness of alternative options but chose the most restrictive one for publishers.
The impact of Google’s AI search features, such as AI Overviews, on publisher traffic has been a growing concern. By providing direct answers to user queries, these features reduce the need for users to click on links, depriving publishers of valuable ad revenue and product sales opportunities. Some publishers have reported significant drops in traffic since the launch of AI Overviews a year ago.
Google’s decision to tie publisher opt-outs for AI features to their overall presence in search results has left many feeling pressured to comply, given the search engine’s overwhelming market share of roughly 90 percent. Publishers rely heavily on Google as a gateway to the modern web, making it difficult for them to opt out of search altogether without suffering severe consequences.
The ongoing antitrust trial against Google has brought these issues to the forefront, with the Department of Justice proposing remedies aimed at curbing the tech giant’s market dominance. One such proposal would compel Google to provide publishers with more granular control over how their content is used in AI models and products, allowing them to opt out on a model-by-model and product-by-product basis without penalty.
Google has defended its position, with Liz Reid, the company’s head of search, testifying that creating multiple opt-outs for different products and models would be challenging and costly. She argued that it would require separate AI models for each feature, adding “enormous complexity” to the system.
Read more at Bloomberg here.
Lucas Nolan is a reporter for Breitbart News covering issues of free speech and online censorship.
Read the full article here