Google is providing new funding to support the DNSRF and GASA in launching the GSE. By combining efforts and creating a centralized platform, GSE seeks to enhance the sharing of abuse signals, facilitating quicker detection and mitigation of fraudulent activities across diverse sectors, platforms and services. The platform’s data engine, hosted on the Google Cloud Platform, enables participants to share and access signals contributed by others. Harnessing Google Cloud’s AI capabilities, it intelligently identifies patterns and matches signals to enhance detection and response efforts. Implementation The intervention is based on three pillars: 1) technical, 2) legal and 3) operational. The DNSRF, a UK-based non-profit organization, initially developed a data engine that runs on Google Cloud Platform to store and analyse DNS-related data for academic and research purposes. Google learned about this platform and DNSRF via its membership in the GASA in early 2024. Google put it to the test, ran a pilot with bad merchant domains from Google Shopping and, in parallel, developed and signed a first-of-its-kind data-sharing agreement with DNSRF. This has allowed Google to test the platform’s capabilities and initiate an internal signal exchange programme to be able to create and manage the exchange of bad actor signals via the GSE. 1) Technical: The intervention is a new global data platform that allows the exchange of bad actor data across sectors and geographies in real time. Participating organizations have the ability to share and receive any type of abuse data to improve their own abuse protections and help detect and prevent scam and fraud campaigns faster.2) Legal: The intervention relies on a data licence agreement that enables GSE to act as a data processor for various types of data, including personal identifiable information of bad actors and most frequently organized crime groups. 3) Operational: This intervention consists of a signal exchange framework through which Google is able to connect its own relevant teams and efforts to the GSE, allowing the exchange to be performed efficiently at scale and driving down the time-to-live, an industry metric indicating the timeframe between when a scam is detected and mitigated. Feedback, measurement and transparency The feedback from users and partners was collected over multiple years, peaking in 2021, indicating that abuse data should be allowed to be shared in order to mitigate the proliferation of scams and fraud. Without a complete picture of the life cycle of a scam, it is hard to pinpoint where it starts, which platforms and services it touches, and which interventions could be effective. In the initial pilot of the data platform, Google was, for the first time, able to share over 100,000 URLs of bad shopping merchants and, as part of the same test, ingest 1 million scam signals. Over time, Google will expand its footprint of platforms and services connected to the GSE, the number of signals shared and ingested as well as actions taken based on the signals exchanged. The impact is measured through identified bad actors across products that, when detected, would be exposed to actions according to the associated policy violations. The Intervention Journey: A Roadmap to Effective Digital Safety Measures 25

The Intervention Journey A Roadmap to Effective Digital Safety Measures 2025