Google's security and privacy managers have warned that certain provisions of the Digital Markets Act (DMA) could lead to increased fraud and the de-anonymization of search queries, according to WIRED, which cites internal documents.

The European Commission is currently preparing final decisions regarding Google Search and Android's compliance with existing legislation. The DMA, passed in late 2022, aims to compel major platforms to open their systems and data to competitors.

In April, the European Commission clarified how Google should share information. Competitors will gain access to data at a level comparable to what the corporation itself uses, including query texts, some metadata, click information, and ranking results.

Heather Adkins, Google's Vice President of Security, told WIRED that the current requirements could negatively impact users in the EU. The company estimates that fraud cases will rise, and if the DMA is enacted, it will create a risk of de-anonymizing queries, making such data sets more attractive targets for hackers.

The European Commission suggests minimizing risks by prohibiting user re-identification and enhancing the security of sensitive information storage. However, Google argues that these measures are insufficient: data will need to be shared with greater detail, and anonymization methods can be compromised by hackers.

Independent security expert Lukasz Olejnik wrote in his blog that data sanitization is inadequate for such a scale. In contrast, DuckDuckGo believes that "the European Commission's approach fits within the existing legal framework," while Alice Cooper, Executive Director of the Knight-Georgetown Institute, described the proposed data-sharing regime as robust and suggested involving independent experts in the review process.

A separate set of requirements pertains to Android. Eugene Liderman, responsible for the operating system's security, warned that malicious actors could exploit broader access to permissions, microphones, cameras, and screen information. According to WIRED, Google's position on this issue is partially supported by Apple.

Previously, Anthropic, known for developing the chatbot Claude, published proposals for the safe deployment of AI agents in corporate environments.