GOOGLE ADAPTS ITS POLICY
The tech giant has changed its policy on abusing a site’s reputation, which had been criticized by publishers. Just to recall, this policy aimed to fight the so-called ‘parasitic SEO.’ This is a practice that involves publishing third-party pages on a well-known site to improve its search ranking.
As the competition watchdog, the European Commission launched an investigation under the DMA (Digital Markets Act), a law that regulates the practices of tech giants. It found that this policy was downgrading websites and news media content in search results when those sites included content from commercial partners. Google said that, starting August 30, the measures to downgrade sites would no longer apply to EU member states, as well as to Iceland, Norway, and Liechtenstein.
CHAT GPT MUST COMPLY TO THE DIGITAL SERVICES ACT
The European Commission has just labeled ChatGPT as a ‘very large online search engine’ (VLOSE) under the Digital Services Act. This classification will subject OpenAI’s service which has about 159 million monthly users in the EU, far surpassing the 45 million threshold to stricter obligations.
THE EU COMMISSION STARTS ENFORCING THE AI ACT
Henna Virkkunen, Executive Vice President of the European Commission in charge of technological sovereignty, security, and democracy, just announced that the European AI Office had sent its first requests for information to several general-purpose model providers “located in different parts of the world.” Their identities haven’t been revealed. Three topics are being looked at: the safety of models, the use of external assessments, and the measures to monitor their behavior once they are on the market. Brussels doesn’t present this approach as just a simple dialogue. According to the vice president, it’s a “first step in implementing the AI Act.”
While the obligations set out by the European text are gradually coming into effect from 2025, a new milestone was reached on August 2, 2026. From that date, the control powers provided by the regulation can be used against providers of AI models for general use. So it’s no longer just about companies saying they’ve put the necessary mechanisms in place. The controllers can now ask them to provide the elements needed to prove it.
At this stage, compliance concerns general-purpose AI models, meaning models that are versatile enough to be used in many systems.
