The European Union Commission has its own places on the function of the Google search in the new “AI” function. Brussels government institutions are currently investigating how “Google AI Overview” works in behavior and how the European Union reporting in fluid texts generated by artificial intelligence, collecting more complex discoveries with the Reporting Law Rules of the European Union, told the magazine Euractiv. According to the Digital Markets Act (DMA) for “goalkeeper”, hard competitive requirements were tightened, according to the Digital Services Act (DSA), other platform requirements and old rules against unfair competition may also be relevant.
So it’s not enough: the commission examines it Report According to the European Union Media Freedom Act, European Media Freedom Act (EMFA), the media is in line with the requirements of AI overview to protect the variety. Among other things, citizens should “reach a large number of editorially independent media materials”. Platform operators are encouraged to ensure this.
Google introduced AI interviews in Germany and other European countries – with some delays, for example, compared to the United States – in late March. Above all, publishers and independent small material manufacturers see it as a problematic that users are often allowed to satisfy themselves with AI interviews and no longer go to the source website for information. With clicks, site operators also lack the possibility of earning money with advertising on their proposal. This traditional online business model AI challenges the discoveries from the Copilot of Chatgpt, Perplexity or Microsoft.
Stringent traffic declines in publication sites
After the introduction of the AI function, many editors have recorded a huge decline in data traffic. Financial service Bloomberg Google gave this effect in a meeting with 20 website manufacturers behind closed doors in October. BBC complains in one Input For the British Government that the AI broker is a threat to his brand. They will not represent the value of reporting public service broadcaster or Article incorrectly summarizeThe system is a well known problem with a generative AI. Google’s AI observation has already been seen by very suspicious summary.
According to Eurachativ, a high-ranking Google manager indicated in early April that the number of discoveries increased after the introduction of AI interviews. This can also benefit publishers. However, the inner formula did not want to reveal any click rates as the number changed continuously.
According to the EU Urleber Rouctive Directive, the tasks protected for research objectives can be used under an exception for lessons and data mining (TDM). “AI -generated summary that reflect preserved materials, TDMs may fall under exceptions,” the Render explained. However, the material should be legally accessible and not subject to the right holder.
No opt-out is possible from AI overview
In theory, author and users who want to stop TDM with their online functions can be reserved for themselves. Such an announcement is only effective when it is “in machine -dialing form”. So usually in the robot. TXT file. However, according to Google, the right holders cannot currently exclude the inclusion in the AI summary-the clause is only related to AI training. According to the American group, the separation between AI search and general discovery will usually lose importance over time.
Google states that rejection of AI training has no effect on the visibility of the respective website in the discovery results. However, the news producers are still afraid that they are slipping into hit lists in such exclusion, Rant Shredeer, director of European Journalist Federation, Eurachave, said. It is committed to a high granulation in the manifestation of the “adequately detailed” summary of training data prescribed in the AI Act, so that rights holders can easily claim compensation for the use of their content. Google advocates in contrast. According to a study commissioning copyright initiative, replication of functions by models for generative AI usually represents a copyright -Relevant reproduction and hence it is illegal.
The Commission found in an initial decision that Google violated DMA as this discovery prefers to treat its own products in engine results. According to Regnier, these initial results do not mention AI overview. They were introduced only after the decision was published. However, similar principles were considered. A week before the introduction of the AI overview in Europe, Google published an experiment that the “no average impact” from the query of removing the material from European news publishers was on the company’s income. This clear performance of Shakti made the publisher angered. According to him, Google benefits from the reliability and reliability of their content, even if the direct income in the group’s news business is relatively low.
(Nen)
