Google Details Who Can License Its Search Data In Europe

Google has revised its documentation for the European Search Dataset Licensing Program, offering elaborate accusation connected really eligible competitors tin entree and licence hunt ranking, query, click, and position information from Search successful the European Economic Area.

The page, past updated connected August 31, covers eligibility requirements, sample datasets, audit procedures, and important deadlines. This inaugural implements the European Commission’s binding decision adopted successful July nether the Digital Markets Act, which mandates Google to stock anonymized hunt information pinch competitors, including AI chatbots that suffice arsenic online hunt engines nether the law.

Licensing agreements will statesman dispatching from September 17, pinch information samples disposable starting November 16.

Who Can Apply

Applicants must meet the criteria of an online hunt motor nether the DMA and target users wrong the EEA. They must not beryllium controlled by non-EEA authorities actors aliases beryllium taxable to EU sanctions.

All applicants must person an mean of astatine slightest 50,000 monthly users of their hunt work successful the EU complete the past year. Additionally, they must person been providing hunt services successful the EU for the past 2 consecutive years or, if they are caller entrants, person received complete 50 cardinal euros successful superior investment.

Google whitethorn petition supporting documents and says it will respond to expressions of liking wrong 7 almanac days.

Data Samples And Key Dates

Eligible applicants whitethorn petition samples earlier purchasing afloat dataset access.

A free sample includes 1,000 rows. A synthetic dataset pinch up to 10 cardinal queries and a 5% sample of the afloat dataset are disposable for a fee.

All 3 samples will beryllium accessible starting November 16.

The fees for the dataset are group connected fair, reasonable, and non-discriminatory terms. According to the Commission’s measures, these fees are constricted to the incremental costs of information proviso positive a specified complaint of return.

Audit Requirements

Gaining entree to the 5% sample and the afloat dataset starts pinch an independent assessment. While the smaller samples don’t require an audit, Google still cautiously checks each applicant’s eligibility successful beforehand to make judge everything is successful place.

An independent auditor cautiously examines the applicant’s method setup, information storage, and workflows. The Level 1 study provides assurance that the applicant has reliable plans to usage the information for its ain hunt product, and that its information protection controls are suitably designed.

To support entree open, recipients work together to ongoing monitoring by an independent assurance practitioner and request to taxable regular Level 2 reports. These reports trial whether the controls are moving successful practice. The first study is owed wrong six months of entree being granted, pinch early reports owed annually.

These audits are important because recipients grip individual data. The Commission believes that keeping the information walled disconnected wrong the recipient’s ain systems tin meet the DMA’s anonymization standard, but it’s up to each applicant to show that this is done properly.

Why This Matters

In July, I discussed the binding decision erstwhile the eligibility thresholds and wide timeline were already public.

The updated page turns the July determination into dates and process. The Level 1 audit gates the 5% sample and the afloat dataset, not the 2 smaller samples. This attack allows eligible companies to analyse the free rows and the synthetic dataset earlier they take to incur immoderate audit costs.

Looking Ahead

Licensing agreements travel first, pinch method specifications connected the dataset to follow.

The Commission plans to reappraisal these measures each 2 years. Plus, Google will request to support a nationalist webpage listing each third-party hunt engines that entree the dataset.

This transparent database is an important measurement to spot if the information extends beyond the big-name engines to see the qualifying chatbots the determination covers.

Featured Image: Net Vector/Shutterstock

Source: Google.