Google does nothing to the authors of the content. Internet creators pay for someone else's innovation
Behind us is the breakthrough phase of the antitrust process of the American Department of Justice versus Google. DEEPMIND Vice President Eli Collins admitted that when the Gemini model passes from the AI laboratory to the search engine, the latter can continue to train it on the content to which the authors forbade access. If the creator wants to protect his text, the only way is to completely turn off the site indexing by the robots.txt file – i.e. Giving up the presence in search results, i.e. most of the internet traffic.
Online content authors So they face a brutal choice: consent to intellectual theft or digital self -ignorance. Naivety costs real money. The emarketer study indicates that “Review from AI” can cut organic traffic by up to 60 percent, and the Raptive advertising network estimates the cumulative revenue gap of the industry at $ 2 billion. annually.
In addition, AI Overviews develop faster than publishers are able to measure their influence. As the interlocutors report Digiday, Google still It does not provide separate statistics of clicks from AI summaries, which prevents thorough analysis and planning of activities.
Google does nothing to the authors of the content
When revenues decrease, the cost of creating value – including research, photos, illustrations – remains unchanged. Meanwhile, the monopolist increases his advantage thanks to the data.
In an internal note, revealed by the Department of Justice, Google filtered “only” 80 billion of 160 billion tokens, whose owners reserved “no -ai” (no access for AI Giant's solutions). The second 80 billion still powers Gemini, and in addition the search engine uses the history of user session and video materials from YouTube.
Google explains that powerful investments in research and development ($ 49 billion in 2024) justify the freedom to use the data, and proposed by the mentioned Department of the Solution – including the order to provide the index with competitors or sell the Chrome browser – threaten “the very existence of” search engines. This line of defense was presented by the CEO of Pichai during the April testimony.
The argument goes to court, but does not solve the problem of asymmetry of strength: Billions of dollars of Google investments come to a large extent from the advertisements displayed next to the content obtained from the same publishers who today are received (read) part of the movement.
Google pays, but the chosen one
Even worse, the Pay -to -Play mechanism works only for the largest. Reddit, having an IPO army, forced Google a license agreement worth around $ 60 million. annually. A single blogger or a niche service of such a gear does not have – However, their texts are still indexed and used without remuneration.
Where's the regulations? These can't keep up. In the European Union, AI ACT, adopted in 2024, imposes an obligation on models of general purpose models … respect for Opt -Out signals and publication of the training collection. Theoretically, the robots.txt system should be enough, but the DSM directive leaves the gate: scraping is legal, Unless the creator reserved the rights in a “universally recognizable” manner. In practice, there is a lack of a uniform standard, and bots can ignore the robots.txt file.
If the courts recognize that the mere summary of someone else's text in AI Overviews does not violate copyright, The creators will be forced to look for new monetization paths: Newsletters, Paywalli, partnerships with AI models or common collective lawsuits. In the USA, the precedent is determined by the New York Times vs OpenAi lawsuit, whose possible victory could break the current rhetoric of technological giants.
A scenario that is not very favorable to democracy and Internet users is also possible. When Google, Otnai and others will collect value, but do not reflect traffic, more and more publishers and authors will start blocking the indexing, move the content for Paywalle and license agreements. Paradoxically, this can accelerate the closing of the open internetwhich was fuel for GPT or Gemini.
What can the authors of the content do now? In fact, they are forced to work. Before them the need to diversify and build direct relations with recipientsbefore the algorithmic source of movement dries completely. Pressure (and hope) rests with the regulators to definitely decide whether aggregation of knowledge by a monopolistic platform without compensation is in the category of “innovation” or already “expropriation” of intellectual value.
